Spectral · Senior ML Researcher (Generative CAD)

completed174 qualified1 runMay 5, 10:08 AMspectral-is-building-generative-foundation-models-that-turn-1777975731

ParsedSpectral · 3 topics · Senior · Hybrid · United States, Canada, or Europe

Generating seed nodes

0 proposed

Explored 0 queries

0/0 done

Expanding nodes

queued

Qualifying candidates

queued

Qualified Candidates (162)

Alexander Raistrick

high hireability

Ph.D. Research Assistant@Princeton University

Previously: Undergraduate Research Assistant - Fouhey AI Lab @ University of Michigan College of Engineering

Princeton, US

3D Geometric Deep Learning55

Generative CAD Models8

CAD Modelling5

Strengths

Infinigen (CVPR 2023, 6.9k stars) — lead author, procedural 3D generation

Infinigen Indoors CVPR 2024 — structured indoor scene synthesis

Gaps

No generative ML model training (diffusion/autoregressive) — core JD requirement

…click to see all

Artem Lukoianov

high hireability

Research Intern@DeepMind

Previously: Research Intern @ Meta

Boston, US

3D Geometric Deep Learning85

Generative CAD Models55

CAD Modelling40

Strengths

Differentiable parametric CAD paper (2024) — directly on-topic

MeshSDF NeurIPS 2020 (196 cit.) — differentiable SDF-to-mesh extraction

Gaps

No evidence of structured/editable CAD output (feature trees, B-rep, STEP)

…click to see all

Arturs Berzins

high hireability

Postdoctoral Researcher@Johannes Kepler Universität Linz

Previously: PhD Fellow @ Johannes Kepler Universität Linz

Linz, AT

3D Geometric Deep Learning72

Generative CAD Models12

CAD Modelling5

Strengths

'Neural Implicit Shape Editing using Boundary Sensitivity' — semantic 3D shape editing

'Shape Generation via Weight Space Learning' (ICLR 2025) — generative shape models

Gaps

No generative CAD experience — no B-rep, STEP, feature trees, or parametric structures

…click to see all

Bing Li

high hireability

Postdoc@King Abdullah University of Science and Technology

Previously: Postdoc @ University of Southern California

3D Geometric Deep Learning68

Generative CAD Models12

CAD Modelling5

Strengths

Magic123 (433 citations) — image-to-3D via 2D+3D diffusion priors

Scene flow in 3D point clouds — geometric deep learning at scale

Gaps

No structured CAD experience — no B-rep, feature tree, STEP, or parametric work

…click to see all

Congyue Deng

high hireability

Student Researcher@DeepMind

Previously: Research Scientist @ Waymo

San Francisco, US

3D Geometric Deep Learning88

Generative CAD Models15

CAD Modelling8

Strengths

Vector Neurons (ICCV 2021 Oral, 430 citations) — seminal equivariant 3D representation

NAP: Neural 3D Articulated Prior — generative prior for structured 3D objects

Gaps

No direct generative CAD experience (B-rep, feature trees, STEP outputs)

…click to see all

David Novotný

high hireability

3D Geometric Deep Learning90

Generative CAD Models20

CAD Modelling5

Strengths

HoloDiffusion (CVPR 2023): 3D diffusion model trained on 2D images

PartGen (2024): part-level structured 3D generation with multi-view diffusion

Gaps

No CAD-specific work: B-rep, STEP, feature trees, or parametric design

…click to see all

Fuyang Zhang

high hireability

Research Assistant@Simon Fraser University

Previously: Intern @ Meta

Burnaby, CA

3D Geometric Deep Learning80

CAD Modelling45

Generative CAD Models35

Strengths

BR-DF (arXiv 2025): B-rep volumetric distance function representation

SceneScript ECCV 2024: autoregressive structured 3D — methodological CAD analog

Gaps

No direct generative CAD work (feature trees, STEP, B-rep generation) — adjacent only

…click to see all

Jianyuan Wang

high hireability

Research Assistant@Meta

London, GB

3D Geometric Deep Learning92

Generative CAD Models20

CAD Modelling3

Strengths

VGGT: CVPR 2025 Best Paper — 3D geometry transformer (234 citations)

AutoPartGen (2025): autoregressive 3D part generation and discovery

Gaps

No CAD/B-rep/STEP/feature-tree or parametric CAD experience found

…click to see all

Lyle Regenwetter

high hireability

Graduate Student@Massachusetts Institute of Technology

Previously: ML Intern, Platform Architecture @ Apple

Boston, US

Generative CAD Models62

3D Geometric Deep Learning48

CAD Modelling42

Strengths

BIKED++: 1.4M bicycle parametric CAD designs — structured generative design at scale

"Constraining Generative Models" — negative data for engineering constraint satisfaction

Gaps

No direct B-rep/STEP/feature-tree work — parametric design, not structured CAD pipelines

…click to see all

Minghao Chen

high hireability

Research Scientist Intern@Meta

Previously: Research Scientist Intern @ Meta

London, GB

3D Geometric Deep Learning87

Generative CAD Models32

CAD Modelling5

Strengths

AutoPartGen (NeurIPS 2025): autoregressive structured 3D part discovery

PartGen (CVPR 2025 Highlight): multi-view diffusion for compositional 3D

Gaps

No direct CAD / B-rep / STEP / feature-tree experience

…click to see all

Mingrui Zhang

high hireability

Ph.D. Candidate@Imperial College London

Previously: Researcher @ Tencent

London, GB

3D Geometric Deep Learning65

Generative CAD Models20

CAD Modelling5

Strengths

PhyCAGE (2024): compositional 3D asset generation from image

BAG3D (2025): body-aligned 3D wearable generation from geometry

Gaps

No structured CAD experience (B-rep, STEP, feature trees, NURBS)

…click to see all

Quankai Gao

high hireability

Ph.D. candidate@University of Southern California

Previously: Research Intern @ Waymo

Los Angeles, US

3D Geometric Deep Learning70

Generative CAD Models8

CAD Modelling5

Strengths

GaussianFlow: 4D content creation via Gaussian dynamics (TMLR, 75 citations)

Can3Tok: 3D tokenization + latent modeling for scene-level Gaussians (ICCV2025)

Gaps

No CAD experience — no B-rep, NURBS, or feature tree work

…click to see all

Sauradip Nag

high hireability

Postdoctoral Researcher@Simon Fraser University

Previously: Research Engineer/Associate @ ICSR and CAIR, DRDO

3D Geometric Deep Learning70

Generative CAD Models5

CAD Modelling3

Strengths

In-2-4D: generative 4D interpolation (SIGGRAPH Asia 2025)

ASIA: few-shot 3D mesh part segmentation (SIGGRAPH Asia 2025)

Gaps

No CAD-specific work — no B-rep, feature trees, STEP, or parametric CAD

…click to see all

Shakiba Kheradmand

high hireability

Research Assistant@University of British Columbia

Previously: Research Intern @ DeepMind

Vancouver, CA

3D Geometric Deep Learning72

CAD Modelling5

Generative CAD Models5

Strengths

NeurIPS 2024 Spotlight: 3DGS-MCMC — core 3D representation ML work

CVPR 2024 Neural Fields paper — implicit 3D representation training

Gaps

No generative CAD experience (B-rep, NURBS, STEP, feature trees)

…click to see all

Yang You

high hireability

Postdoctoral Researcher@Stanford University

Previously: PhD @ Shanghai Jiao Tong University

Stanford, US

3D Geometric Deep Learning82

Generative CAD Models60

CAD Modelling42

Strengths

Img2CAD (SIGGRAPH Asia 2025): structured CAD output from images via VLM-assisted conditional factorization

SPRIN/PRIN: rotation-invariant point cloud features (TPAMI 2021)

Gaps

Img2CAD is reverse engineering (image→CAD), not forward generative CAD synthesis

…click to see all

Aditya Balu

medium hireability

Data Scientist@Translational AI Center (TrAC), Iowa State University

Previously: Postdoctoral Research Associate @ Iowa State University

Ames, US

3D Geometric Deep Learning80

CAD Modelling60

Generative CAD Models52

Strengths

NURBSDiff: differentiable NURBS module — core CAD geometry ML (2022, 46 cit.)

Latent diffusion for structural component design — generative 3D (2024)

Gaps

No feature-tree or B-rep generative CAD output — focus is engineering geometry

…click to see all

Aditya Sanghi

medium hireability

Director@Samruddhi

Previously: Director @ Sanghi Industries Ltd

Ahmedabad, IN

Generative CAD Models95

3D Geometric Deep Learning92

CAD Modelling78

Strengths

SolidGen (2023): autoregressive B-rep synthesis — Spectral's exact domain

BRepNet (177 citations): topological message passing on solid CAD geometry

Gaps

MS only, no PhD — research depth vs. academic competition

…click to see all

Ajad Chhatkuli

medium hireability

Researcher@INSAIT

Previously: Senior Scientist @ Computer Vision Lab, ETH Zurich

3D Geometric Deep Learning75

Generative CAD Models5

CAD Modelling3

Strengths

Neural Vector Fields (NeurIPS 2022) — implicit 3D surface representation

7 years at ETH Zurich CVL — production-grade 3D geometric DL research

Gaps

Zero generative CAD work — no B-rep, STEP, feature tree, or editable CAD outputs

…click to see all

Alexey Artemov

medium hireability

Computer Vision Research Engineer@Apple

Previously: Senior Research Scientist @ Technical University of Munich

Munich, DE

3D Geometric Deep Learning83

Generative CAD Models48

CAD Modelling42

Strengths

ABC dataset (CVPR 2019, 691 citations) — landmark CAD dataset for geometric DL

MeshGPT (2024, 203 citations) — generative mesh via autoregressive transformer

Gaps

MeshGPT generates triangle meshes, not structured CAD (B-rep/STEP/feature trees)

…click to see all

Aliasghar Khani

medium hireability

Researcher@Autodesk

Previously: AI Research Internship @ Autodesk

Vancouver, CA

3D Geometric Deep Learning65

Generative CAD Models20

CAD Modelling15

Strengths

WaLa: billion-parameter 3D diffusion model on implicit surfaces (SDF)

Autodesk Research — working in 3D generative domain

Gaps

No structured CAD generation (B-rep, feature trees, STEP) in any work

…click to see all

Aljaz Bozic

medium hireability

Research Scientist@Meta

Previously: Research Scientist @ Meta

Zurich, CH

3D Geometric Deep Learning75

Generative CAD Models10

CAD Modelling5

Strengths

NPMs: Neural Parametric Models (133 cites) — parametric 3D shape generation

Neural Deformation Graphs CVPR'21 — 3D geometric deep learning

Gaps

No CAD-specific work — no B-rep, feature trees, STEP, or parametric CAD

…click to see all

Amin Heyrani Nobari

medium hireability

Research Assistant@MIT

Previously: Research Assistant @ University of Toronto

Boston, US

CAD Modelling65

Generative CAD Models58

3D Geometric Deep Learning55

Strengths

Cad-coder (2025): VLM generating CAD code from images — direct generative CAD

cadeval2 active contributor Jan 2026 — structured CAD evaluation

Gaps

No published work on B-rep, feature trees, or STEP-format structured output

…click to see all

Amir Hertz

medium hireability

Research Scientist@Google

Previously: Researcher @ Google

3D Geometric Deep Learning85

Generative CAD Models20

CAD Modelling5

Strengths

SPAGHETTI: part-aware implicit 3D shape generation/editing (SIGGRAPH 2022)

MeshCNN: edge-based 3D mesh CNN, 954 citations (SIGGRAPH 2019)

Gaps

No CAD-specific work: no B-rep, STEP, feature trees, or parametric CAD

…click to see all

Amlan Kar

medium hireability

Senior Research Scientist@NVIDIA

Previously: Research Scientist @ NVIDIA

Toronto, CA

3D Geometric Deep Learning48

Generative CAD Models22

CAD Modelling8

Strengths

ATISS: autoregressive transformers for 3D scene synthesis (NeurIPS 2021, 215 citations)

Meta-Sim: structured scene graph generation for synthetic datasets (ICCV 2019, 331 citations)

Gaps

No parametric CAD work — no B-rep, STEP, or feature tree generation

…click to see all

Andrei Zanfir

medium hireability

Researcher@Google

Previously: PhD student @ Institute of Mathematics of the Romanian Academy

3D Geometric Deep Learning72

Generative CAD Models10

CAD Modelling3

Strengths

GHUM: generative 3D human shape model (470 citations, 2020)

DreamHuman: text→3D avatar via NeRF + score distillation (125 cit.)

Gaps

Zero CAD domain knowledge: no B-rep, feature trees, STEP, or NURBS work

…click to see all

Anqi Pang

medium hireability

ShanghaiTech University

Previously: Researcher @ PCG

Shanghai, CN

3D Geometric Deep Learning82

Generative CAD Models45

CAD Modelling5

Strengths

MeshXL: autoregressive mesh generation, same paradigm as structured CAD generation

CLAY (2024): 239-citation 3D generative model published in ACM TOG

Gaps

No CAD-specific work: B-rep, STEP, feature trees absent from publication record

…click to see all

Arianna Rampini

medium hireability

Senior Research Scientist@Autodesk

Previously: Junior Research Scientist @ Autodesk

3D Geometric Deep Learning82

Generative CAD Models25

CAD Modelling20

Strengths

Make-A-Shape (ICML 2024): 10M-scale 3D generative model — proven foundation model builder

WaLa: billion-parameter 3D diffusion with wavelet encodings — large-scale 3D generation

Gaps

No B-rep, feature-tree, or STEP output work — models produce general 3D shapes, not structured CAD

…click to see all

AU

medium hireability

CAD Modelling92

3D Geometric Deep Learning25

Generative CAD Models5

Strengths

863 commits to CadQuery/cadquery — primary maintainer by large margin

NURBS MVP (Apr 2026) — implements B-rep surface/curve evaluation from scratch

Gaps

No ML or generative model work found anywhere in profile

…click to see all

Baorui Ma

medium hireability

Researcher@Beijing Academy of Artificial Intelligence

Previously: PhD student @ Tsinghua University

Beijing, CN

3D Geometric Deep Learning88

Generative CAD Models20

CAD Modelling5

Strengths

See3D (CVPR 2025 Highlight): generative 3D creation from video at scale

UDiFF: diffusion model for 3D unsigned distance fields (CVPR 2024)

Gaps

No generative CAD experience: no feature trees, B-rep, STEP, or structured parametric outputs

…click to see all

Basile Van Hoorick

medium hireability

Research Scientist@Toyota Research Institute

Previously: Graduate Research Assistant @ Columbia University

San Francisco, US

3D Geometric Deep Learning72

Generative CAD Models8

CAD Modelling3

Strengths

Zero-1-to-3 (ICCV 2023, 1328 cites) — pioneered single-image 3D generation

Generative Camera Dolly: ECCV 2024 Oral, extreme novel view synthesis

Gaps

No CAD or parametric geometry experience (no B-rep, STEP, feature trees)

…click to see all

Beichen Li

medium hireability

Member of Technical Staff@OpenAI

Previously: Graduate Research Assistant @ MIT

San Francisco, US

3D Geometric Deep Learning30

Generative CAD Models22

CAD Modelling8

Strengths

VLMaterial (ICLR 2025 Spotlight) — LLM → structured procedural node graph generation

RL for procedural material generation (SIGGRAPH Asia 2024)

Gaps

No 3D shape representation work (B-rep, mesh, point cloud, NURBS)

…click to see all

Benjamin Eckart

medium hireability

CTO (Contract – Independent Consultant)@Benifex

Previously: Acting CTO (SVP Platform & Enablement) @ Babbel

Southampton, GB

3D Geometric Deep Learning78

Generative CAD Models10

CAD Modelling5

Strengths

DeepGMR (333 citations): generative latent models for 3D point cloud registration

HGMR: hierarchical Gaussian mixtures for adaptive 3D representations (146 citations)

Gaps

No CAD/B-rep/NURBS/STEP or feature tree experience

…click to see all

BinbinHuang

medium hireability

PhD Student@The University of Hong Kong

Previously: PhD Student @ ShanghaiTech University

Hong Kong, HK

3D Geometric Deep Learning82

Generative CAD Models10

CAD Modelling3

Strengths

2DGS (SIGGRAPH 2024, 976 citations): geometrically accurate radiance fields

Mip-Splatting: CVPR 2024 Oral + Best Student Paper, 728 citations

Gaps

Zero CAD experience: no B-rep, NURBS, feature trees, STEP, or parametric design

…click to see all

Boxiao Pan

medium hireability

Research Scientist@Luma AI

Previously: PhD student @ Stanford University

3D Geometric Deep Learning72

Generative CAD Models18

CAD Modelling3

Strengths

EG3D (1.8K citations, CVPR 2022 Oral) — co-authored foundational 3D generative model

PartNeRF — part-aware editable 3D shapes (editability mirrors CAD design concepts)

Gaps

No CAD/B-rep/feature-tree work — all 3D work uses implicit neural representations

…click to see all

Ceyuan Yang

medium hireability

Research Scientist@Shanghai Artificial Intelligence Laboratory

Previously: Researcher @ ByteDance

San Francisco, US

3D Geometric Deep Learning62

Generative CAD Models5

CAD Modelling2

Strengths

GRM: Gaussian reconstruction model for 3D generation (234 citations, 2024)

BerfScene (CVPR 2024): infinite 3D scene generation via radiance fields

Gaps

No B-rep, feature trees, STEP, or parametric CAD experience

…click to see all

Chao Liu

medium hireability

Researcher@NVIDIA

Previously: Research And Development Scientist @ Elephas

Madison, US

3D Geometric Deep Learning65

Generative CAD Models12

CAD Modelling5

Strengths

BlobGEN-3D: compositional 3D-consistent freeview generation (NVIDIA GenAIR, 2024)

3D point cloud generation — self-supervised discrete generative models (83 citations)

Gaps

No CAD-specific work — no B-rep, STEP, feature trees, or parametric geometry

…click to see all

Chengcheng Tang

medium hireability

Research Scientist@Meta

Previously: Postdoc @ Stanford University

3D Geometric Deep Learning65

CAD Modelling55

Generative CAD Models20

Strengths

PhD under Helmut Pottmann (KAUST) — leading computational geometry lineage

DeepSpline: data-driven parametric curve/surface reconstruction (2019)

Gaps

No evidence of generative models for structured CAD (B-rep/STEP/feature trees)

…click to see all

Chen Wang

medium hireability

PhD student@University of Pennsylvania

Previously: MS student @ Tsinghua University

Philadelphia, US

3D Geometric Deep Learning78

Generative CAD Models12

CAD Modelling5

Strengths

tttLRM (CVPR 2026 Highlight): autoregressive Gaussian splatting 3D reconstruction

threestudio co-developer: unified 3D generation framework (206 citations)

Gaps

No CAD-specific experience: no B-rep, STEP, feature trees, or parametric CAD

…click to see all

Clemens Eppner

medium hireability

Research Scientist@NVIDIA

Previously: Senior Research Scientist @ NVIDIA

Seattle, US

3D Geometric Deep Learning62

Generative CAD Models12

CAD Modelling5

Strengths

6-DOF GraspNet (VAE for 3D grasp generation, 786 citations)

DiMSam: diffusion models for structured task/motion planning

Gaps

Zero CAD experience — no B-rep, STEP, NURBS, or feature tree work

…click to see all

David McAllister

medium hireability

Research Scientist Intern@NVIDIA

Previously: Research Scientist Intern @ Luma AI

Helsinki, FI

3D Geometric Deep Learning45

Generative CAD Models5

CAD Modelling0

Strengths

Flow Matching Policy Gradients (ICLR 2026) — RL post-training for generative models

Nerfstudio co-author (881 citations) — 3D neural rendering/implicit surfaces

Gaps

No CAD, B-rep, STEP, or structured/parametric geometry work found

…click to see all

Daxuan Ren

medium hireability

Principal Software Development Engineer@Autodesk

Previously: Algorithm Researcher @ SenseTime

Singapore, SG

Generative CAD Models92

3D Geometric Deep Learning82

CAD Modelling80

Strengths

PartCAD (NeurIPS 2025): autoregressive CAD sequence generation from point clouds

CVPR 2026 paper: parametric CAD sketch generation from bidirectional query

Gaps

Singapore-based; may require relocation or remote arrangement

…click to see all

Despoina Paschalidou

medium hireability

Senior Research Scientist@NVIDIA

Previously: Postdoctoral Researcher @ Stanford University

San Francisco, US

3D Geometric Deep Learning88

Generative CAD Models20

CAD Modelling5

Strengths

ATISS: autoregressive transformer for 3D scene synthesis (NeurIPS 2021)

Neural Parts + PartNeRF: learning editable 3D shape abstractions

Gaps

No structured CAD work — no B-rep, STEP, or feature tree evidence

…click to see all

Edith Tretschk

medium hireability

Research Scientist@Meta

Previously: PhD Candidate @ Max Planck Institute for Informatics

San Francisco, US

3D Geometric Deep Learning72

CAD Modelling5

Generative CAD Models5

Strengths

PatchNets: patch-based deep implicit 3D shape representations (126 citations)

DEMEA: mesh autoencoder for non-rigid deforming objects — 3D geometry ML

Gaps

No generative model papers — work is reconstruction not generation

…click to see all

Edward J. Smith

medium hireability

Machine Learning Researcher@Borealis AI

Previously: Visiting Researcher @ Meta

Toronto, CA

3D Geometric Deep Learning82

Generative CAD Models15

CAD Modelling5

Strengths

Kaolin co-author — NVIDIA PyTorch 3D DL library (118 citations)

GEOMetrics (ICML 2019): mesh-graph generative models for 3D

Gaps

No CAD domain experience — no B-rep, feature trees, STEP, or parametric structure

…click to see all

Fabian Manhardt

medium hireability

Research Scientist@Google

Previously: PhD Student @ Technical University Munich

Bavaria, DE

3D Geometric Deep Learning75

Generative CAD Models8

CAD Modelling5

Strengths

TextMesh: text-to-3D mesh generation via diffusion (2024, 150 citations)

NerfMeshing: NeRF to geometrically-accurate 3D meshes (2024, 85 citations)

Gaps

No CAD-specific work — no B-rep, STEP, feature trees, or parametric design

…click to see all

Fangyin Wei

medium hireability

Research Scientist@NVIDIA

Previously: Research Intern @ Meta

San Francisco, US

3D Geometric Deep Learning83

Generative CAD Models15

CAD Modelling5

Strengths

Edify 3D (2024): NVIDIA production 3D asset generation — scalable generative pipeline

PartPacker (NeurIPS 2025): part-level 3D generation with structured decomposition

Gaps

No CAD-specific work: no B-rep, STEP, feature trees, or parametric modeling

…click to see all

Fenggen Yu

medium hireability

Applied Scientist@Amazon

Previously: Applied Scientist II @ Amazon

San Francisco, US

3D Geometric Deep Learning88

CAD Modelling80

Generative CAD Models78

Strengths

PhD thesis: Learning Structured Representations of 3D CAD Models (SFU, 2024)

CAPRI-Net: generative CAD via primitive assembly, 92 citations

Gaps

Current role at Amazon Prime Video (image/video) — drifting from CAD research

…click to see all

Feng Liu

medium hireability

Principal Research Scientist@Adobe

Previously: Consultant @ Meta

Portland, US

3D Geometric Deep Learning58

Generative CAD Models10

CAD Modelling5

Strengths

LRM (ICLR 2024): single-image to 3D, transformer-based large reconstruction model

Progressive Autoregressive Video Diffusion (ICLR 2025): autoregressive generative architecture

Gaps

No CAD work: no B-rep, STEP, feature-tree, or parametric geometry experience

…click to see all

Filippos Kokkinos

medium hireability

Research Scientist@Meta

Previously: PhD Candidate @ University College London

London, GB

3D Geometric Deep Learning82

Generative CAD Models12

CAD Modelling5

Strengths

VFusion3D (2024, 56 cit) — scalable 3D generative model from video diffusion

IM-3D (2024, 76 cit) — multiview diffusion for high-quality 3D generation

Gaps

No structured CAD experience (B-rep, STEP, feature trees, parametric models)

…click to see all

Francis Williams

medium hireability

Senior Research Scientist@NVIDIA

Previously: Student Researcher @ NVIDIA

New York, US

3D Geometric Deep Learning95

Generative CAD Models45

CAD Modelling20

Strengths

XCube (CVPR 2024 highlight) — large-scale 3D generative model on sparse voxels

fVDB creator — production 3D deep learning library at NVIDIA scale

Gaps

No structured CAD generation — outputs are point clouds/voxels, not B-rep or feature trees

…click to see all

Giorgio Giannone

medium hireability

Principal Research Scientist@Red Hat

Previously: Applied Scientist @ Amazon

Boston, US

Generative CAD Models82

CAD Modelling68

3D Geometric Deep Learning62

Strengths

GIFT (ICML 2026): image-to-CAD program synthesis via geometric feedback

Co-led Text2CAD: NL → structured CAD generative model at MIT DeCoDE Lab

Gaps

No explicit B-rep / STEP / feature-tree generative model work found

…click to see all

Gopal Sharma

medium hireability

Senior Researcher@Samsung

Previously: Advisor @ AuraML

San Francisco, US

3D Geometric Deep Learning92

Generative CAD Models80

CAD Modelling52

Strengths

CSGNet: neural CSG program parser — structured output directly analogous to CAD feature trees

ParSeNet: NURBS/parametric surface fitting on 3D point clouds — B-rep adjacent

Gaps

No explicit B-rep, STEP, or feature tree generation work — CSG is adjacent but not identical to parametric CAD

…click to see all

Guocheng Gordon Qian

medium hireability

Research Scientist@Snap

Previously: Research Intern @ Snap

San Francisco, US

3D Geometric Deep Learning88

Generative CAD Models18

CAD Modelling5

Strengths

PointNeXt (NeurIPS 2022, 1027+ cites) — state-of-art 3D point cloud learning

Magic123 (ICLR 2024, 432+ cites) — image-to-3D diffusion generative pipeline

Gaps

No structured CAD experience: no B-rep, NURBS, STEP, or parametric feature trees

…click to see all

Haitao Yang

medium hireability

Ph.D. student@University of Texas at Austin

Previously: Research intern @ Waymo

Austin, US

3D Geometric Deep Learning78

Generative CAD Models15

CAD Modelling5

Strengths

Atlas Gaussians Diffusion (ICLR 2025 Spotlight) — diffusion over 3D Gaussian representations

CoFie (NeurIPS 2024) — compact neural implicit surface fields

Gaps

No CAD domain experience: no B-rep, feature trees, STEP, or parametric structures

…click to see all

Hanlin Goh

medium hireability

Independent Researcher@Sabbatical

Previously: Machine Learning Researcher @ Apple

San Francisco, US

3D Geometric Deep Learning50

Generative CAD Models10

CAD Modelling5

Strengths

GAUDI (NeurIPS 2022): generative 3D scene model at Apple scale

Geometric Capsule Autoencoders: 3D point cloud representation learning

Gaps

No CAD-specific work: no B-rep, feature trees, or parametric structures

…click to see all

Hansheng Chen

medium hireability

Intern@Adobe

Previously: Research Assistant @ Northwestern University

San Francisco, US

3D Geometric Deep Learning78

Generative CAD Models40

CAD Modelling20

Strengths

Img2CAD (2025): VLM-based 3D CAD reverse engineering from images

Zero123++ (2023, 398 citations): multi-view 3D generation, widely used

Gaps

No evidence of structured/editable CAD generation (feature trees, B-rep, parametric output)

…click to see all

Hanxue Liang

medium hireability

PhD student@University of Cambridge

Previously: Research Scientist Intern @ NVIDIA

3D Geometric Deep Learning58

Generative CAD Models5

CAD Modelling3

Strengths

L4GM: first 4D large reconstruction model, NeurIPS 2024 (82 cites)

Diffusion4D: diffusion-based 4D generation with Gaussian splatting (64 cites)

Gaps

No CAD-specific work — zero B-rep, NURBS, feature trees, or parametric structure

…click to see all

Hao Tan

medium hireability

Research Scientist@Adobe

Previously: Research Intern @ Bloomberg

San Francisco, US

3D Geometric Deep Learning82

Generative CAD Models8

CAD Modelling5

Strengths

LRM (620 citations) — foundational large 3D reconstruction model at Adobe

MeshLRM (2025) — high-quality mesh generation, production scale

Gaps

No structured/parametric CAD work — all 3D is neural mesh/gaussian reconstruction

…click to see all

Hsueh-Ti Derek Liu

medium hireability

Senior Research Scientist@Roblox

Previously: Research Scientist @ Roblox

Vancouver, CA

3D Geometric Deep Learning88

Generative CAD Models42

CAD Modelling22

Strengths

Octree-based autoregressive shape generation (2025) -- novel tokenization for 3D

CSG on neural SDFs (2023) -- boolean generative geometry akin to CAD paradigm

Gaps

No B-rep, STEP, or parametric feature-tree CAD experience

…click to see all

Jan Eric Lenssen

medium hireability

Senior Researcher / Group Leader@Max Planck Institute for Informatics

Previously: Postdoctoral Researcher @ Max Planck Institute for Informatics

Saarbrücken, DE

3D Geometric Deep Learning88

Generative CAD Models25

CAD Modelling8

Strengths

SplineCNN: B-Spline conv kernels for geometric DL (641 cites, CVPR 2018)

Deep Local Shapes: SDF priors for 3D reconstruction (567 cites, ECCV 2020)

Gaps

No CAD-specific work: B-rep, NURBS, STEP, or feature tree experience

…click to see all

Jiacheng Chen

medium hireability

PhD Candidate@Simon Fraser University

Previously: Ph.D. student @ University of Southern California

Vancouver, CA

3D Geometric Deep Learning52

Generative CAD Models30

CAD Modelling12

Strengths

PolyDiffuse: diffusion for structured polygon shape output (NeurIPS 2023)

Floor-SP: Inverse CAD for floorplans, sequential structured prediction (ICCV 2019)

Gaps

No B-rep, NURBS, STEP, or feature-tree CAD experience

…click to see all

Jiahui Huang

medium hireability

Senior Research Scientist@NVIDIA

Previously: Research Scientist @ NVIDIA

San Francisco, US

3D Geometric Deep Learning92

Generative CAD Models18

CAD Modelling8

Strengths

XCube (CVPR 2024, 133 cit.): generative 3D via sparse voxel hierarchies

DiffFacto (ICCV 2023): structured part-based 3D generation with diffusion

Gaps

No B-rep, feature tree, or STEP file generation work found

…click to see all

Jiaxiang Tang

medium hireability

Research Scientist@NVIDIA

Previously: Intern @ NVIDIA

San Francisco, US

3D Geometric Deep Learning88

Generative CAD Models22

CAD Modelling5

Strengths

DreamGaussian (ICLR 2024 Oral, 905 cites) — core 3D generative model author

MeshAnything — autoregressive transformer for artist-quality mesh generation

Gaps

No structured/parametric CAD experience (B-rep, STEP, feature trees)

…click to see all

Jing Li

medium hireability

PhD student@University of Science and Technology of China

Generative CAD Models90

3D Geometric Deep Learning82

CAD Modelling62

Strengths

DTGBrepGen (CVPR 2025) — first-authored B-rep generative model for CAD

B-spline topology/geometry decoupling — exact SGS-1/SGS-2 design space

Gaps

Junior researcher — h-index 3, still PhD student level

…click to see all

Jingwei Huang

medium hireability

Ph.D Candidate@Stanford University

Previously: Principal Researcher @ Tencent

San Francisco, US

3D Geometric Deep Learning87

Generative CAD Models32

CAD Modelling18

Strengths

Hunyuan3D-2 contributor — large-scale generative 3D diffusion at Tencent

TextureNet (CVPR 2019 Oral) — learning from high-resolution signals on meshes

Gaps

No structured/parametric CAD work (B-rep, feature trees, STEP, NURBS)

…click to see all

Jonathan Masci

medium hireability

Principal Researcher@NNAISENSE

Previously: Postdoc @ Università della Svizzera italiana

3D Geometric Deep Learning85

Generative CAD Models5

CAD Modelling3

Strengths

MoNet (2525 citations) — foundational spectral GDL on graphs and manifolds

Geodesic CNNs (978 citations) — pioneer of 3D shape convolutions

Gaps

No generative CAD work — no feature tree, B-rep, STEP, or structured output generation

…click to see all

Julian Chibane

medium hireability

PhD Student@Max Planck Institute for Informatics

Saarbrücken, DE

3D Geometric Deep Learning78

CAD Modelling5

Generative CAD Models5

Strengths

NDF (NeurIPS 2020, 407 cites) — canonical implicit 3D field method

IF-Net (CVPR 2020) — feature-space implicit function for shape completion

Gaps

Zero generative CAD experience — no B-rep, STEP, or feature-tree work

…click to see all

Junzhe Zhang

medium hireability

Senior Staff Engineer@Huawei

Previously: Algorithm Researcher @ SenseTime

Singapore, SG

3D Geometric Deep Learning76

CAD Modelling55

Generative CAD Models48

Strengths

ExtrudeNet: sketch-and-extrude shape parsing — CAD-native ML representation

CSG-Stump: CSG-tree structured shape understanding (ICCV 2021, 65 cites)

Gaps

No forward generative CAD — ExtrudeNet/CSG-Stump are inverse/parsing, not generation

…click to see all

Kamal Rahimi Malekshan

medium hireability

Principal Machine Learning Engineer@Autodesk

Previously: Senior Research Engineer - Machine Learning @ Autodesk

Toronto, CA

Generative CAD Models92

3D Geometric Deep Learning90

CAD Modelling82

Strengths

Zero-to-CAD (2025): CAD program synthesis at million-scale, no real data

Editable prismatic CAD from voxels (SIGGRAPH Asia 2022) — structured, editable output

Gaps

No public RL-for-generative-models work (RLHF/DPO/GRPO)

…click to see all

Karsten Kreis

medium hireability

Principal Research Scientist@NVIDIA

Previously: Senior Research Scientist II @ NVIDIA

Vancouver, CA

3D Geometric Deep Learning82

Generative CAD Models15

CAD Modelling5

Strengths

LION: latent point diffusion for 3D shape generation (NeurIPS 2022, 634 citations)

Magic3D text-to-3D (CVPR 2023, 1461 citations); L4GM, Align Your Gaussians

Gaps

No structured CAD experience — no B-rep, STEP, feature trees, or parametric design work

…click to see all

Kseniya Cherenkova

medium hireability

RnD Software Engineer@Artec 3D

Previously: PhD Student @ University of Luxembourg

Luxembourg, LU

CAD Modelling92

3D Geometric Deep Learning88

Generative CAD Models75

Strengths

CAD-Recode (2025): generates executable Python CAD code from point clouds

TransCAD (2024): hierarchical transformer for CAD sequence inference, SOTA on DeepCAD/Fusion360

Gaps

Primarily reconstruction/reverse engineering; less evidence of unconditional generative models (diffusion/AR)

…click to see all

Lars Mescheder

medium hireability

3D Geometric Deep Learning92

Generative CAD Models15

CAD Modelling5

Strengths

Occupancy Networks (NeurIPS 2019): primary author of foundational implicit 3D rep paper

Convolutional Occupancy Networks — scales 3D reconstruction to complex scenes

Gaps

No CAD-specific work — no B-rep, STEP, feature tree, or parametric CAD experience

…click to see all

Lawson Fulton

medium hireability

3D Geometric Deep Learning70

Generative CAD Models62

CAD Modelling55

Strengths

Augmenta founding engineer -- generative 3D design for automated MEP/BIM

Augmenta builds structured AI outputs for construction CAD at scale

Gaps

No evidence of B-rep/STEP/feature-tree generation specifically

…click to see all

Maham Tanveer

medium hireability

Ph.D. student@Simon Fraser University

Previously: Software Engineer @ Sedenius Technologies

3D Geometric Deep Learning68

Generative CAD Models60

CAD Modelling42

Strengths

D²CSG (NeurIPS 2023): generative CSG tree learning for 3D CAD shapes

Unsupervised learning of compact Boolean CSG representations — structured output

Gaps

Recent work (2024-2025) pivoted to video inbetweening, not 3D CAD

…click to see all

Marco Fumero

medium hireability

Postdoctoral Researcher@Institute of Science and Technology Austria

Previously: Doctoral Researcher @ Sapienza Università di Roma

Vienna, AT

3D Geometric Deep Learning72

Generative CAD Models12

CAD Modelling5

Strengths

CLIP-Forge (CVPR 2022): zero-shot text-to-3D generation with neural implicits

Nonlinear spectral geometry processing — ACM Trans. Graph. 2020, meshes + point clouds

Gaps

No CAD-specific work: no B-rep, STEP, parametric structure, or feature tree experience

…click to see all

Maria Shugrina

medium hireability

3D Geometric Deep Learning75

Generative CAD Models18

CAD Modelling8

Strengths

ATISS (NeurIPS 2021): autoregressive transformer for generative 3D scenes

3DStyleNet (ICCV 2021): ML-driven geometric+texture 3D shape generation

Gaps

No CAD modeling experience — no B-rep, STEP, feature trees, or parametric CAD

…click to see all

Marie-Julie Rakotosaona

medium hireability

PhD student@Ecole Polytechnique

3D Geometric Deep Learning85

Generative CAD Models12

CAD Modelling5

Strengths

Learning Delaunay Surface Elements — CVPR 2021 oral, ML-driven mesh reconstruction

Differentiable Surface Triangulation — differentiable 3D geometry generation (SIGGRAPH Asia 2021)

Gaps

No generative CAD work: no structured/editable output, no B-rep or feature tree generation

…click to see all

Matheus Gadelha

medium hireability

Research Scientist@Adobe

Previously: Research Assistant @ University of Massachusetts Amherst

Seattle, US

3D Geometric Deep Learning90

Generative CAD Models35

CAD Modelling10

Strengths

GEM3D (SIGGRAPH 2024): generative medial abstractions for 3D shape synthesis

SuperFrusta (CVPR 2026): primitive fitting on 3D shapes — structured decomposition

Gaps

No B-rep, STEP, or parametric CAD feature-tree work found

…click to see all

Matthew Fisher

medium hireability

Principal Scientist@Adobe

Previously: Graduate student @ Stanford University

3D Geometric Deep Learning82

Generative CAD Models18

CAD Modelling10

Strengths

AtlasNet (CVPR 2018): pioneering 3D surface generation — cited 1000+

ShapeShifter (CVPR 2025): 3D diffusion on point-voxel representations

Gaps

No structured/parametric CAD work (B-rep, STEP, feature trees) in publications

…click to see all

Michael Oechsle

medium hireability

Research Scientist@Google

Previously: Researcher @ ETAS

Zurich, CH

3D Geometric Deep Learning85

Generative CAD Models15

CAD Modelling5

Strengths

Occupancy Networks (3684 cites, CVPR 2019) — foundational implicit 3D learning

UNISURF (ICCV 2021 oral) — neural surface reconstruction, key 3D geometry work

Gaps

No CAD-specific experience — no B-rep, STEP, feature trees, or parametric CAD

…click to see all

Mikaela Angelina Uy

medium hireability

Researcher@NVIDIA

Previously: PhD Student @ Stanford University

San Francisco, US

3D Geometric Deep Learning92

Generative CAD Models78

CAD Modelling72

Strengths

Img2CAD (2024): VLM-assisted 3D CAD reverse engineering from images

Point2Cyl (CVPR 2022, 91 citations): point cloud → extrusion cylinders

Gaps

No B-rep or STEP-level structured CAD output — work is point-cloud/image-to-cylinder, not full feature-tree generation

…click to see all

Nicholas Sharp

medium hireability

Senior Research Scientist@NVIDIA

Previously: Postdoctoral Research Fellow @ University of Toronto

Seattle, US

3D Geometric Deep Learning90

Generative CAD Models18

CAD Modelling5

Strengths

SpaceMesh: continuous representation for generating valid manifold meshes (SIGGRAPH Asia 2024)

DiffusionNet: 326-citation landmark paper for ML on 3D surfaces

Gaps

No CAD-specific work: no B-rep, STEP, feature trees, or parametric CAD generation

…click to see all

Nishkrit Desai

medium hireability

Researcher@Axiom

Previously: Intern @ NVIDIA

Generative CAD Models92

3D Geometric Deep Learning72

CAD Modelling65

Strengths

SolidGen: autoregressive B-rep synthesis model (TMLR/ICLR 2024, 89 citations)

Autodesk research: transformer generative model for CAD solids

Gaps

Junior profile — h-index 3, limited publication depth

…click to see all

Olga Diamanti

medium hireability

3D Geometric Deep Learning80

CAD Modelling45

Generative CAD Models35

Strengths

ICML 2018: representation learning + GANs for 3D point clouds (957 citations)

Autodesk Research principal AI researcher — industry ML for 3D geometry

Gaps

No generative parametric CAD (B-rep/STEP/feature tree) work found

…click to see all

Or Perel

medium hireability

Research Scientist@NVIDIA

Previously: Research Scientist @ Amazon

3D Geometric Deep Learning82

CAD Modelling45

Generative CAD Models25

Strengths

SPAGHETTI (SIGGRAPH 2022): part-aware generative 3D shape editing

Kaolin contributor (20 commits) — NVIDIA's core 3D DL library

Gaps

No structured/parametric CAD (B-rep, feature trees, STEP) — works in neural implicit space

…click to see all

Pavel Tokmakov

medium hireability

Machine Learning Researcher@Toyota

Previously: Postdoc @ CMU

San Francisco, US

3D Geometric Deep Learning78

Generative CAD Models15

CAD Modelling5

Strengths

Zero-1-to-3: image-to-3D via diffusion model (1323 citations, ICCV 2023)

Generative 4D Gaussian Splatting — dynamic scene generation from video (2025)

Gaps

No CAD-specific work — no B-rep, feature trees, STEP, or parametric geometry generation

…click to see all

Philipp Henzler

medium hireability

Senior Research Scientist@Google

Previously: Research Scientist @ Google

San Francisco, US

3D Geometric Deep Learning68

Generative CAD Models10

CAD Modelling5

Strengths

Bolt3D + CAT3D — generative 3D diffusion models at Google (2024-2025)

UFO-4D (ICLR 2026): feedforward 4D reconstruction — recent top-venue work

Gaps

No CAD-specific work — all 3D work is visual/neural rendering, not parametric/B-rep/STEP

…click to see all

Pim De Haan

medium hireability

Member of technical staff@CuspAI

Previously: Senior engineer @ Qualcomm

3D Geometric Deep Learning82

Generative CAD Models12

CAD Modelling3

Strengths

Geometric Algebra Transformers (NeurIPS 2023) — architecture for 3D geometric data

Gauge equivariant mesh CNNs — ML directly on 3D mesh geometry

Gaps

No CAD experience — all 3D work is materials science, not engineering geometry

…click to see all

Pradeep Kumar Jayaraman

medium hireability

Generative CAD Models100

CAD Modelling95

3D Geometric Deep Learning95

Strengths

BrepGen: B-rep generative diffusion model (SIGGRAPH 2024)

SolidGen: autoregressive direct B-rep synthesis (TMLR)

Gaps

~6 years at Autodesk — long tenure, no 'open to work' signals

…click to see all

Pradyumna Reddy

medium hireability

Senior Research Scientist@Autodesk

Previously: Senior Research Scientist @ Huawei

London, GB

3D Geometric Deep Learning92

CAD Modelling72

Generative CAD Models70

Strengths

DualBRep (SIGGRAPH 2026): deep learning + B-rep CAD representation, exact target domain

WaLa (ICLR 2025): billion-parameter 3D generative diffusion model, Autodesk AI Lab

Gaps

3D generative work uses implicit SDF/wavelet fields, not structured CAD feature trees or STEP

…click to see all

Qimin Chen

medium hireability

PhD student@Simon Fraser University

Previously: Research Scientist/Engineer Intern @ Adobe

Burnaby, CA

3D Geometric Deep Learning78

Generative CAD Models55

CAD Modelling30

Strengths

DCSG NeurIPS 2023 — generative CSG tree model (structured 3D, 52 citations)

UNIST CVPR 2022 — neural implicit shape translation

Gaps

No directly attributed B-rep, STEP, or feature-tree CAD generation work

…click to see all

Robert Giaquinto

medium hireability

Principal AI Research Scientist, Manager@Autodesk

Previously: Applied Scientist @ Amazon

Los Angeles, US

Generative CAD Models55

CAD Modelling45

3D Geometric Deep Learning30

Strengths

Autodesk AI Lab — building 3D foundation models for AEC domain (generative CAD adjacent)

PhD in deep generative models (normalizing flows, VAEs) from U Minnesota

Gaps

No public 3D geometric deep learning papers — B-rep, NURBS, point clouds not demonstrated

…click to see all

Ruiqi Gao

medium hireability

Staff Research Scientist@DeepMind

Previously: Research Scientist @ Google

San Francisco, US

3D Geometric Deep Learning78

Generative CAD Models18

CAD Modelling5

Strengths

CAT3D: multi-view diffusion for 3D creation (CVPR/NeurIPS 2024)

Bolt3D: fast 3D scene generation, Google DeepMind 2025

Gaps

No parametric CAD experience — works on 3D scenes/objects, not B-rep/feature trees

…click to see all

Ruiyu Wang

medium hireability

PhD student@University of Toronto

Previously: Research Assistant @ Microsoft

Toronto, CA

Generative CAD Models85

CAD Modelling55

3D Geometric Deep Learning25

Strengths

CADFusion ICML 2025 first-author — text-to-CAD with visual feedback (DPO)

microsoft/CADFusion: official impl, 77 stars, production-quality codebase

Gaps

No explicit 3D geometry work — approach is LLM sequence-based, not mesh/B-rep/NURBS

…click to see all

Sai Bi

medium hireability

Senior Research Scientist@Adobe

Previously: Research Intern @ Meta

San Francisco, US

3D Geometric Deep Learning85

Generative CAD Models15

CAD Modelling5

Strengths

LRM (ICLR 2024 oral, 620 cites) — pioneered single-image-to-3D large reconstruction models

DMV3D: multi-view diffusion + 3D LRM (ICLR 2024 spotlight, 205 cites)

Gaps

No generative CAD work — no B-rep, STEP, feature tree, or parametric structure modeling

…click to see all

Sam Xu (Xiang Xu)

medium hireability

Generative CAD Models99

3D Geometric Deep Learning95

CAD Modelling90

Strengths

BrepGen (SIGGRAPH 2024): diffusion model for B-rep generation

DualBrep (SIGGRAPH 2026): dual-field continuous B-rep representation

Gaps

At Autodesk Research — may have competing non-compete constraints

…click to see all

Sergey Zakharov

medium hireability

Senior Research Scientist@Toyota Research Institute

Previously: Research Scientist @ Toyota Research Institute

San Francisco, US

3D Geometric Deep Learning78

Generative CAD Models12

CAD Modelling3

Strengths

Zero-1-to-3 (ICCV 2023) — zero-shot diffusion model for image-to-3D generation

ROAD (CoRL 2022) — implicit recursive octree auto-decoder for 3D shapes

Gaps

No CAD-domain experience: no B-rep, STEP, feature trees, or parametric structure

…click to see all

Songyou Peng

medium hireability

Senior Research Scientist@DeepMind

Previously: Senior Researcher/PostDoc @ ETH Zürich

San Francisco, US

3D Geometric Deep Learning88

Generative CAD Models12

CAD Modelling3

Strengths

ConvONet (ECCV 2020, 1278 citations) — landmark implicit surface paper

Shape As Points (NeurIPS 2021) — differentiable mesh reconstruction

Gaps

Zero generative CAD experience — no B-rep, feature trees, STEP, or parametric outputs

…click to see all

Souhaib Attaiki

medium hireability

Research Scientist@Veeton

Previously: PhD Candidate @ École Polytechnique

Paris, FR

3D Geometric Deep Learning88

Generative CAD Models20

CAD Modelling8

Strengths

DiffusionNet (SIGGRAPH 2022, 326 citations) — discretization-agnostic surface ML

GANFusion: text-to-3D with diffusion from Adobe Research internship

Gaps

No CAD-specific work — no B-rep, feature trees, STEP, or NURBS experience

…click to see all

Stefan Popov

medium hireability

Staff AI Research Scientist@Meta

Previously: Staff Software Engineer @ DeepMind

Zurich, CH

3D Geometric Deep Learning70

CAD Modelling20

Generative CAD Models12

Strengths

C-flow (CVPR 2020): generative flow models for 3D point clouds

CoReNet (ECCV 2020): coherent 3D scene reconstruction, 80 citations

Gaps

No structured/editable CAD generation — no B-rep, feature tree, or STEP output work

…click to see all

Stephan J. Garbin

medium hireability

Senior Research Scientist@Google

Previously: Senior Researcher @ Microsoft

London, GB

3D Geometric Deep Learning72

Generative CAD Models3

CAD Modelling2

Strengths

FastNeRF (798 citations) — neural rendering, implicit 3D representations

Binary Opacity Grids (2024) — fine geometric detail for mesh-based synthesis

Gaps

No CAD modeling experience — no B-rep, feature trees, or STEP file work

…click to see all

Tolga Birdal

medium hireability

Assistant Professor@Imperial College London

Previously: Postdoc @ Stanford University

London, GB

3D Geometric Deep Learning88

CAD Modelling32

Generative CAD Models28

Strengths

Point2Cyl (98 cit): point cloud → extrusion cylinders, structured parametric reconstruction

PPFNet (813 cit) + PPF-FoldNet (506 cit) — top-cited 3D point ML

Gaps

No direct generative CAD work — no feature tree, B-rep, or STEP generation papers

…click to see all

Tom Monnier

medium hireability

Research Scientist@Meta

Previously: Research intern @ Adobe

3D Geometric Deep Learning88

Generative CAD Models30

CAD Modelling5

Strengths

PartGen (CVPR 2025 Highlight) — compositional part-level 3D generation via diffusion

AutoPartGen (NeurIPS 2025) — autoregressive 3D part discovery and generation

Gaps

No CAD-specific work — B-rep, feature trees, STEP, parametric structure absent

…click to see all

Vikram Voleti

medium hireability

Research Scientist@Stability AI

Previously: Research Intern @ Meta

Kitchener, CA

3D Geometric Deep Learning78

Generative CAD Models15

CAD Modelling5

Strengths

SV3D (ECCV Oral 2024): single-image 3D generation via video diffusion

SV4D: dynamic 3D generation with multi-view/multi-frame consistency

Gaps

No structured/parametric CAD experience (B-rep, feature trees, STEP, NURBS)

…click to see all

Wang Yifan

medium hireability

Researcher@Adobe

Previously: Postdoc @ Stanford University

3D Geometric Deep Learning82

Generative CAD Models8

CAD Modelling3

Strengths

RigAnything (2025): autoregressive structured 3D generation at Adobe

GRM (2024): large-scale Gaussian 3D reconstruction/generation model

Gaps

Zero CAD work: no B-rep, STEP, feature trees, or parametric modelling

…click to see all

Weikai Chen

medium hireability

Principal Research Scientist@Tencent

3D Geometric Deep Learning82

Generative CAD Models25

CAD Modelling15

Strengths

NerVE: parametric curve extraction from 3D point clouds (CVPR 2023)

Light-SQ: structured superquadric shape abstraction for generated meshes (SIGGRAPH Asia 2025)

Gaps

No explicit CAD domain knowledge — no B-rep, STEP, or feature-tree work

…click to see all

Weiwei Sun

medium hireability

Applied Scientist@Amazon

Previously: Scientist intern @ Amazon

Vancouver, CA

3D Geometric Deep Learning75

Generative CAD Models15

CAD Modelling10

Strengths

3DGS as MCMC (NeurIPS'24 Spotlight) — generative probabilistic 3D approach

Polygonal geometry representation learning (55 cites) — closest to structured 3D

Gaps

No generative CAD work — no B-rep, STEP, feature trees, or parametric structures

…click to see all

Xin Tong

medium hireability

Partner Research Manager@Anuttacon

Previously: Partner Research Manager @ Microsoft

San Francisco, US

3D Geometric Deep Learning95

Generative CAD Models70

CAD Modelling62

Strengths

ComplexGen: B-rep chain complex generation for CAD (2022, 126 cites)

Structured 3D Latents: scalable generative 3D models (2025, 297 cites)

Gaps

No explicit feature-tree or parametric/procedural CAD generation work

…click to see all

Xin Yu

medium hireability

Research Scientist Intern@Adobe

Previously: Research Scientist Intern @ Adobe

San Francisco, US

3D Geometric Deep Learning78

Generative CAD Models8

CAD Modelling5

Strengths

TEXGen (SIGGRAPH Asia 2024, Best Paper HM) — diffusion model for 3D mesh textures

Text-to-3D Classifier Score Distillation — ICLR 2024, 120 citations

Gaps

No CAD-specific work — no B-rep, STEP, feature tree, or parametric CAD experience

…click to see all

Yanir Kleiman

medium hireability

Research Engineer@Meta

Previously: Postdoc @ École Polytechnique

3D Geometric Deep Learning80

Generative CAD Models20

CAD Modelling5

Strengths

Meta 3D AssetGen: text-to-mesh + PBR materials at scale (2024)

Meta 3D TextureGen: fast consistent 3D texture generation (2024)

Gaps

No structured CAD output — all 3D work is mesh/texture, not B-rep or feature trees

…click to see all

Yaroslav Ganin

medium hireability

Member of Technical Staff@OpenAI

Previously: Co-Founder @ Udio

London, GB

Generative CAD Models82

3D Geometric Deep Learning75

CAD Modelling60

Strengths

"Computer-Aided Design as Language" (NeurIPS 2021) — ML-based CAD sketch generation, first author

PolyGen (ICML 2020) — autoregressive 3D mesh model with Transformers

Gaps

No evidence of B-rep/feature-tree/STEP CAD (SGS-1/SGS-2 style outputs)

…click to see all

Yawar Siddiqui

medium hireability

Research Scientist@Meta

Previously: AI Research Scientist Intern @ Meta

Munich, DE

3D Geometric Deep Learning88

Generative CAD Models20

CAD Modelling5

Strengths

MeshGPT (CVPR24 Highlight) — autoregressive transformer for mesh generation

PolyDiff + VertexRegen — diffusion & LOD-based mesh generation

Gaps

No structured/parametric CAD experience — all work is polygonal mesh, not B-rep or STEP

…click to see all

Yizhi Wang

medium hireability

Research Scientist@ByteDance

Previously: Postdoc Fellow @ Simon Fraser University

3D Geometric Deep Learning78

Generative CAD Models35

CAD Modelling28

Strengths

SweepNet (ECCV 2024): shape abstraction via CAD-style sweep surfaces

GALA (ICLR 2025): 3D generative model with geometry-aware grid representations

Gaps

No direct B-rep, STEP, or parametric feature-tree work visible

…click to see all

Yuan-Chen Guo

medium hireability

Researcher@VAST

Previously: PhD student @ Tsinghua University

3D Geometric Deep Learning92

Generative CAD Models8

CAD Modelling5

Strengths

Threestudio: primary author (249 commits), 7K-star unified 3D generation framework

TripoSG co-author: large-scale rectified flow transformer for 3D shape synthesis (2025)

Gaps

No CAD-specific work: zero papers on B-rep, STEP, feature trees, or parametric structures

…click to see all

Zhengqing Wang

medium hireability

Ph.D. student@Simon Fraser University

Montreal, CA

Generative CAD Models90

3D Geometric Deep Learning78

CAD Modelling30

Strengths

BrepGen (SIGGRAPH 2024, 76 citations) — generative diffusion for B-rep CAD

PuzzleFusion++ (ICLR 2025) — 3D fracture assembly via diffusion

Gaps

Early PhD (year 1-2) — full-time departure from PhD uncertain

…click to see all

Zhiqin Chen

medium hireability

3D Geometric Deep Learning92

Generative CAD Models50

CAD Modelling40

Strengths

CAPRI-Net: primitive assembly for compact CAD shapes (CVPR 2022)

BSP-Net: CSG-style structured 3D generation, Best Student Paper CVPR 2020

Gaps

No published work on B-rep, STEP, or feature-tree structured CAD

…click to see all

Zhongshi Jiang

medium hireability

3D Geometric Deep Learning82

CAD Modelling30

Generative CAD Models15

Strengths

Surface Networks CVPR 2018 Oral — ML/VAE generative model on 3D meshes

ABC dataset co-author — 1M parametric CAD models for geometric DL

Gaps

No evidence of generative CAD outputs (feature trees, B-rep, STEP files)

…click to see all

Zian Wang

medium hireability

Research Scientist@NVIDIA

Previously: Research Scientist Intern @ NVIDIA

Toronto, CA

3D Geometric Deep Learning78

Generative CAD Models28

CAD Modelling5

Strengths

GET3D: generative 3D textured mesh model (NeurIPS 2022, 595 citations)

Flexible Isosurface Extraction: gradient-based mesh optimization

Gaps

No structured CAD experience: B-rep, feature trees, STEP, parametric modelling

…click to see all

Zike Wu

medium hireability

PhD student@University of British Columbia

ex-Nanyang Technological University

3D Geometric Deep Learning62

Generative CAD Models8

CAD Modelling3

Strengths

Gamba (ECCV 2024): 3DGS + Mamba single-view reconstruction, 59 citations

MVGamba: SSM-based 3D content generation, 15 citations

Gaps

No CAD-specific work: no B-rep, STEP, NURBS, feature-tree, or parametric models

…click to see all

Alex Nichol

low hireability

3D Geometric Deep Learning92

Generative CAD Models45

CAD Modelling30

Strengths

Point-E: text→3D point cloud diffusion model — first-author, shipped at OpenAI (2022)

Shap-E: generative model for 3D implicit functions (meshes + NeRF) — first-author (2023)

Gaps

No evidence of parametric/structured CAD (B-rep, STEP, feature trees) — only general 3D geometry

…click to see all

Ali Mahdavi Amiri

low hireability

Vice President Research@Monsters Aliens Robots Zombies

Previously: Research Director @ Monsters Aliens Robots Zombies

North Vancouver, CA

3D Geometric Deep Learning88

Generative CAD Models75

CAD Modelling65

Strengths

CAPRI-Net: ML for compact CAD via adaptive primitive assembly (107 cit.)

D²CSG: learns CSG trees from 3D shapes — structured CAD output (62 cit.)

Gaps

No feature-tree or B-rep/STEP generative model work (CSG/primitives only)

…click to see all

Andrea Tagliasacchi

low hireability

Associate Professor (Status Only)@University of Toronto

Previously: Staff Research Scientist @ DeepMind

Burnaby, CA

3D Geometric Deep Learning90

Generative CAD Models20

CAD Modelling8

Strengths

BSP-Net (381 citations): generative structured mesh via binary space partitioning

CvxNet (307 citations): learnable convex decomposition of 3D shapes

Gaps

No B-rep, feature tree, or STEP file generation work — not generative CAD

…click to see all

Ari Seff

low hireability

Research Scientist@OpenAI

Previously: Research Scientist @ Apple

Generative CAD Models90

CAD Modelling70

3D Geometric Deep Learning50

Strengths

Vitruvion (ICLR 2022): autoregressive generative model of parametric CAD sketches

SketchGraphs: 15M-sketch CAD dataset with geometric constraint graphs

Gaps

Work is primarily 2D CAD sketches — not full 3D B-rep/feature trees

…click to see all

Carlos Esteves

low hireability

Research Scientist@Google

Previously: PhD Candidate @ University of Pennsylvania

New York, US

3D Geometric Deep Learning88

Generative CAD Models22

CAD Modelling5

Strengths

Single Mesh Diffusion Models (CVPR 2024) — generative 3D mesh diffusion model

Spherical CNNs series — equivariant SO(3) 3D representations, 653 citations

Gaps

No CAD/B-rep/STEP/NURBS or structured parametric geometry experience

…click to see all

Cengiz Oztireli

low hireability

Researcher@Google

Cambridge, GB

3D Geometric Deep Learning85

CAD Modelling5

Generative CAD Models5

Strengths

FrePolad (ECCV 2024): VAE+DDPM latent diffusion for point cloud generation

Differentiable Surface Splatting — 375 citations, foundational 3D geometry DL

Gaps

No CAD-specific work: no B-rep, feature trees, STEP, or parametric geometry

…click to see all

Changjian Li

low hireability

Assistant Professor@University of Edinburgh

Previously: Starting Researcher @ Inria

Edinburgh, GB

Generative CAD Models92

CAD Modelling90

3D Geometric Deep Learning72

Strengths

Free2CAD (SIGGRAPH 2022, 122 citations) — freehand drawings → CAD commands

Sketch2CAD (SIGGRAPH Asia 2020, 106 citations) — sequential CAD generation

Gaps

Assistant professor actively building lab — deeply committed to academic track

…click to see all

Chang Xiao

low hireability

Assistant Professor of Computer Science@Boston University

Previously: Research Scientist @ Adobe

Boston, US

Generative CAD Models82

CAD Modelling55

3D Geometric Deep Learning35

Strengths

DeepCAD (ICCV 2021, 302 citations) — seminal transformer for CAD op sequences

Designed dataset of 178K CAD models with construction sequences

Gaps

No follow-up CAD generative work post-2021 — research pivoted to HCI/AR

…click to see all

Charles Loop

low hireability

3D Geometric Deep Learning88

CAD Modelling68

Generative CAD Models10

Strengths

Inventor of Loop subdivision surfaces — 2551-citation foundational 3D algorithm

NGLOD (2022, 607 cit) — neural implicit 3D geometry at SIGGRAPH scale

Gaps

No generative CAD output work — reconstruction & rendering focus, not structured generation

…click to see all

Charlie Nash

low hireability

Research Scientist@OpenAI

Previously: Co-Founder @ Udio

London, GB

3D Geometric Deep Learning82

Generative CAD Models35

CAD Modelling5

Strengths

PolyGen (NeurIPS 2020, 348 citations): autoregressive 3D mesh generation

Shape VAE (2017): generative model of part-segmented 3D objects

Gaps

No CAD-specific work: B-rep, NURBS, STEP, feature trees absent from profile

…click to see all

Cheng Lin

low hireability

Senior Site Reliability Engineer@Alibaba Cloud

Previously: Site Reliability Engineer @ Tencent

Seattle, US

3D Geometric Deep Learning88

Generative CAD Models65

CAD Modelling45

Strengths

CADDreamer (CVPR 2025) — CAD generation from images, direct query match

3DModelingRL (ECCV 2020) — RL for 3D shape modeling, 57 citations

Gaps

No evidence of structured CAD (feature trees, B-rep, STEP) — shape generation, not parametric CAD

…click to see all

Christian Laforte

low hireability

3D Geometric Deep Learning72

CAD Modelling28

Generative CAD Models20

Strengths

VP Research at Stability AI — led 3D GenAI team

13+ commits to threestudio (Zero123, text-to-3D diffusion)

Gaps

No evidence of structured/parametric CAD (B-rep, STEP, feature tree) generative models

…click to see all

Difan Liu

low hireability

Research Scientist@Adobe

Previously: PhD student @ University of Massachusetts, Amherst

3D Geometric Deep Learning82

Generative CAD Models70

CAD Modelling55

Strengths

CSGNet (CVPR 2018, 279 citations) — generates CSG programs from 3D shapes

ParSeNet (ECCV 2020, 214 citations) — parametric surface fitting from point clouds

Gaps

Recent work (2024-2025) pivoted to video/image generation — CAD no longer central

…click to see all

Gang YU

low hireability

Principal Research Scientist@StepFun

Previously: Research Scientist @ Tencent

Rowland Heights, US

3D Geometric Deep Learning92

Generative CAD Models22

CAD Modelling5

Strengths

MeshAnything (ICLR 2025): autoregressive mesh generation — closest to structured 3D output

MeshXL (NeurIPS 2024): neural coordinate fields for generative 3D foundation models

Gaps

No evidence of CAD-specific work (B-rep, STEP, feature trees, parametric CAD)

…click to see all

Hao Pan

low hireability

Assistant Professor@Tsinghua University

Previously: Researcher @ Microsoft

Beijing, CN

CAD Modelling90

Generative CAD Models87

3D Geometric Deep Learning83

Strengths

ComplexGen (SIGGRAPH 2022): B-rep chain complex generative model, 126 citations

Free2CAD + Sketch2CAD: CAD command parsing/generation series

Gaps

New professor (Sep 2024) — actively building lab, not job-seeking

…click to see all

Hooman Shayani

low hireability

Principal Researcher@Autodesk

London, GB

3D Geometric Deep Learning95

CAD Modelling85

Generative CAD Models65

Strengths

BRepNet (CVPR 2021, 159 citations) — pioneered topological ML on solid B-rep

UV-Net (CVPR 2021, 125 citations) — ML directly on boundary representations

Gaps

No evidence of structured/editable CAD output generation (feature trees, STEP)

…click to see all

Janne Hellsten

low hireability

3D Geometric Deep Learning52

Generative CAD Models18

CAD Modelling4

Strengths

StyleGAN2 + StyleGAN3 co-author — core NVIDIA generative model team

nvdiffrast paper: differentiable rasterization of 3D meshes in PyTorch

Gaps

No structured CAD experience — B-rep, NURBS, feature trees entirely absent

…click to see all

Joe Lambourne

low hireability

Generative CAD Models99

CAD Modelling98

3D Geometric Deep Learning97

Strengths

BrepGen (SIGGRAPH 2024) — diffusion-based B-rep generative model, co-author

SolidGen (TMLR 2023) — autoregressive direct B-rep synthesis

Gaps

15+ years at Autodesk — deeply entrenched, hireability likely low

…click to see all

Jonathan Lorraine

low hireability

Research Scientist@NVIDIA

Previously: Research Scientist @ Google

Toronto, CA

3D Geometric Deep Learning80

Generative CAD Models22

CAD Modelling5

Strengths

LLaMA-Mesh: 3D mesh generation with language models (NVIDIA 2024)

LATTE3D: large-scale text-to-3D, 400ms inference (NVIDIA 2024)

Gaps

No CAD-specific work — meshes/NeRF, not B-rep, feature trees, or STEP

…click to see all

Jon Hasselgren

low hireability

Principal Research Scientist@NVIDIA

Previously: Senior Research Scientist @ NVIDIA

Lund, SE

3D Geometric Deep Learning88

Generative CAD Models25

CAD Modelling5

Strengths

FlexiCubes — novel isosurface rep for gradient-based mesh optimization

Edify 3D — shipped scalable generative 3D asset model at NVIDIA

Gaps

No CAD-specific work: B-rep, STEP, feature trees absent from publication record

…click to see all

Jun Gao

low hireability

Research Scientist@NVIDIA

Previously: Research Intern @ NVIDIA

Toronto, CA

3D Geometric Deep Learning92

Generative CAD Models15

CAD Modelling8

Strengths

GET3D (NeurIPS 2022): lead author, generative 3D textured shape model

DMTet (NeurIPS 2021): hybrid representation for high-res 3D shape synthesis

Gaps

No structured CAD work: no B-rep, STEP, feature tree, or parametric CAD models

…click to see all

Ka-Hei Hui

low hireability

Researcher@Autodesk

Previously: PhD student @ The Chinese University of Hong Kong

3D Geometric Deep Learning85

CAD Modelling35

Generative CAD Models30

Strengths

Make-A-Shape (ICML 2024) — first author, 10M-scale 3D shape diffusion model

Neural Wavelet-domain Diffusion — strong wavelet-based 3D generative work

Gaps

Published work on implicit/SDF/wavelet reps — not parametric CAD (B-rep, feature trees, STEP)

…click to see all

Karl D.D. Willis

low hireability

Generative CAD Models99

CAD Modelling95

3D Geometric Deep Learning95

Strengths

BrepGen (SIGGRAPH 2024) — B-rep generative diffusion model, directly SGS-1-style work

SkexGen (ICML 2022) — autoregressive CAD construction sequences, ICML venue

Gaps

Senior Research Manager role — management track, not pure IC researcher

…click to see all

Lars Morten Mescheder

low hireability

Apple

Previously: Graduate Student @ Max Planck Institute for Intelligent Systems

3D Geometric Deep Learning92

Generative CAD Models20

CAD Modelling5

Strengths

Occupancy Networks (CVPR 2019): foundational implicit 3D paper, 4041 citations

Differentiable Volumetric Rendering (CVPR 2020): implicit 3D without 3D supervision

Gaps

No CAD-specific work: no B-rep, STEP, feature trees, or structured 3D generation

…click to see all

Moritz Ibing

low hireability

Senior Machine Learning Engineer@Stealth Startup

Previously: Machine Learning Researcher @ Dataful Minds

Cologne, DE

3D Geometric Deep Learning88

Generative CAD Models22

CAD Modelling10

Strengths

Octree Transformer: autoregressive 3D shape generation with transformers (2021)

CVPR 2021: 3D generative model with grid-based implicit functions

Gaps

No CAD-specific work: no B-rep, feature trees, STEP, or parametric structures

…click to see all

Nick Richardson

low hireability

Senior Engineer@Arm

Previously: Technical Consultant @ Ventrilo.ai

Boston, US

Generative CAD Models82

CAD Modelling62

3D Geometric Deep Learning22

Strengths

Vitruvion (ICLR 2022): generative parametric CAD sketches, 100 citations

Parametric sketch constraints — deep CAD structure understanding

Gaps

No 3D geometric DL (Vitruvion is 2D sketch, not point clouds/meshes/B-rep)

…click to see all

Nigel J. W. Morris

low hireability

Autodesk

Previously: Researcher @ Autodesk

Generative CAD Models85

CAD Modelling78

3D Geometric Deep Learning60

Strengths

SolidGen (TMLR 2023): autoregressive B-rep synthesis — direct query match

Transformer+pointer network on B-rep vertex/edge/face hierarchy

Gaps

No evident diffusion or RL-for-generative-models (RLHF/DPO) experience

…click to see all

Niloy Mitra

low hireability

Principal Scientist@Adobe

Previously: Researcher @ Adobe

London, GB

Generative CAD Models95

3D Geometric Deep Learning90

CAD Modelling88

Strengths

SketchGen (NeurIPS 2021, 109 cit) — generative constrained CAD sketches

Free2CAD (2022, 122 cit) — freehand drawing to CAD command generation

Gaps

Tenured Full Professor at UCL — extremely low probability of recruiting

…click to see all

Paul Guerrero

low hireability

Research Scientist@Adobe

Previously: Research Associate (Post-Doc) @ University College London

London, GB

3D Geometric Deep Learning85

Generative CAD Models75

CAD Modelling55

Strengths

SketchGen (NeurIPS 2021): constrained CAD sketch generation model

ShapeMOD: macro operation discovery for 3D shape programs

Gaps

No evidence of B-rep, STEP, or feature-tree level CAD generation

…click to see all

Peter Wonka

low hireability

Researcher@Snap

Previously: Full Professor @ King Abdullah University of Science and Technology

Thuwal, SA

3D Geometric Deep Learning88

CAD Modelling68

Generative CAD Models65

Strengths

SketchGen (2021): generates constrained CAD sketches — direct match

PS-CAD (2025): CAD reconstruction with generative guidance

Gaps

No demonstrated B-rep / feature-tree / parametric CAD generation (closer to geometry reconstruction)

…click to see all

Rundi Wu

low hireability

Research Scientist@DeepMind

Previously: PHD Student @ Columbia University

San Francisco, US

Generative CAD Models95

3D Geometric Deep Learning90

CAD Modelling65

Strengths

DeepCAD (ICCV 2021): created foundational generative CAD model

178K CAD model dataset — built the benchmark data for the field

Gaps

Only ~11 months at DeepMind — low hireability window

…click to see all

Shizhao Sun

low hireability

Principal Researcher@Microsoft

Previously: PhD student @ Nankai University

Generative CAD Models92

CAD Modelling72

3D Geometric Deep Learning15

Strengths

FlexCAD (ICLR 2025): controllable CAD generation via LLM fine-tuning

Text-to-CAD (ICML 2025): visual feedback loop for CAD generation quality

Gaps

3D geometry focus is weak — works with CAD token sequences, not B-rep/NURBS/mesh directly

…click to see all

Siddhartha Chaudhuri

low hireability

Senior Research Scientist@Adobe

Previously: Assistant Professor @ IIT Bombay

New York, US

3D Geometric Deep Learning92

Generative CAD Models55

CAD Modelling35

Strengths

GRASS (2017, 487 cit.): generative structured 3D shape synthesis

ParSeNet (2020, 214 cit.): parametric surface fitting from 3D point clouds

Gaps

No explicit CAD output work (B-rep, STEP, feature trees) in publication record

…click to see all

S. M. Ali Eslami

low hireability

Distinguished Research Scientist@DeepMind

Previously: Principal Research Scientist @ DeepMind

London, GB

3D Geometric Deep Learning82

Generative CAD Models45

CAD Modelling5

Strengths

PolyGen: autoregressive Transformer for 3D polygon mesh generation (ICML 2020)

GQN: neural 3D scene representation and rendering, Science 2018

Gaps

No CAD-specific work — no B-rep, feature trees, STEP, or parametric structures

…click to see all

Thibault Groueix

low hireability

Research Scientist@Meta

Previously: Research Scientist @ Adobe

San Francisco, US

3D Geometric Deep Learning92

Generative CAD Models15

CAD Modelling8

Strengths

AtlasNet: CVPR 2018 spotlight, 1,678 citations — foundational 3D generative model

Meta RS: large-scale diffusion models for 3D generative AI

Gaps

No generative CAD experience — no B-rep, feature trees, STEP, or parametric outputs

…click to see all

Tianchang (Frank) Shen

low hireability

3D Geometric Deep Learning88

Generative CAD Models15

CAD Modelling5

Strengths

FlexiCubes (ACM ToG 2023): gradient-based differentiable mesh optimization

GEN3C (CVPR 2025 highlight): 3D-consistent video generation, co-first author

Gaps

No CAD-specific experience — meshes/video, not B-rep/feature trees/STEP

…click to see all

Vladimir Kim

low hireability

Director of AI/ML@Adobe

Previously: Postdoc @ Stanford University

3D Geometric Deep Learning92

Generative CAD Models75

CAD Modelling65

Strengths

"LLM-Driven CAD Design" (2024) — solver-aided DSL for generative CAD

GEM3D (2024): topology-aware 3D diffusion using medial axis transforms

Gaps

No evidence of B-rep or structured feature-tree CAD generation specifically

…click to see all

Xiaohui Zeng

low hireability

Research Scientist@NVIDIA

Previously: PhD student @ University of Toronto

San Francisco, US

3D Geometric Deep Learning92

Generative CAD Models30

CAD Modelling5

Strengths

LION (642 citations) — latent diffusion over 3D point clouds, nv-tlabs

GET3D co-author (NeurIPS 2022) — generative textured 3D mesh model

Gaps

No structured CAD work — all outputs are meshes/voxels/point clouds, not B-rep or feature trees

…click to see all

Yasutaka Furukawa

low hireability

Principal Scientist@Wayve

Previously: Associate Professor @ Simon Fraser University

Generative CAD Models97

3D Geometric Deep Learning90

CAD Modelling80

Strengths

BrepGen: B-rep generative diffusion model (SIGGRAPH 2024) — PI

SkexGen: autoregressive CAD construction sequences (ICML 2022) — author

Gaps

Tenured professor at SFU — very hard to recruit full-time

…click to see all

Yewen Pu

low hireability

Assistant Professor@Nanyang Technological University

Previously: Principal Research Scientist @ Autodesk

CAD Modelling78

Generative CAD Models65

3D Geometric Deep Learning35

Strengths

CadVLM: generative parametric CAD sketch model (2024)

InverseCSG: 3D models → CSG trees, structured CAD recovery

Gaps

CAD sketch-level, not full B-rep/STEP/feature-tree generative models

…click to see all

Yingtian Liu

low hireability

3D Geometric Deep Learning80

Generative CAD Models20

CAD Modelling5

Strengths

FACE (2026): autoregressive mesh generation — same paradigm as SGS-1

NeuFrameQ (ICCV 2025): deep mesh topology/quadrangulation expertise

Gaps

No CAD-specific work — no B-rep, STEP, feature tree or parametric CAD experience

…click to see all

Zan Gojcic

low hireability

Senior Research Manager@NVIDIA

Previously: Research Scientist Intern @ NVIDIA

Zurich, CH

3D Geometric Deep Learning90

Generative CAD Models30

CAD Modelling8

Strengths

GET3D (NeurIPS 2022, 595 cites) — generative 3D textured mesh model

LION (NeurIPS 2022, 637 cites) — latent diffusion for 3D shape generation

Gaps

No B-rep, STEP, feature tree, or parametric CAD experience found

…click to see all

Zixin Zou

low hireability

3D Geometric Deep Learning88

Generative CAD Models20

CAD Modelling5

Strengths

TripoSG: 1.5B param rectified flow model for image-to-3D mesh (2025)

SparseFlex: 1024³ resolution arbitrary-topology mesh modeling

Gaps

No structured parametric CAD work — outputs are meshes/SDFs, not B-rep/feature trees

…click to see all

Runs

#1completed0 qualified / 0 foundMay 5, 10:08 AM