Back to dashboard

# Research Scientist - Synthetic Data ## **What we're looking for** Spectral i…

completed43 qualified1 runApr 29, 8:36 PMresearch-scientist-synthetic-data-what-were-looking-for-spe-1777495007
ParsedSpectral · 4 topics · Mid · Hybrid · San Francisco, US
Generating seed nodes
0 proposed
Explored 0 queries
0/0 done
    3
    Expanding nodes
    queued
    4
    Qualifying candidates
    queued

    Qualified Candidates (40)

    DC

    Di Chang

    high hireability

    Student Researcher@Meta

    Previously: Research Scientist Intern @ Meta

    San Francisco, US

    • Final-year USC PhD (2022-2026) with strong 3D generation and multi-view geometry background: RC-MVSNet (ECCV 2022, 87 citations), MagicPose4D (articulated 3D generation, TMLR 2025), X-Dyna (CVPR 2025 Highlight, diffusion-based human video generation)
    • Work spans 3D reconstruction, diffusion-based video/image generation, and generative modeling — directly matching JD's 'other 3D domains' and 'image/video/world modeling' criteria
    • Currently Research Scientist Intern at Meta SuperIntelligence Labs in Palo Alto (Jan-Aug 2026), SF area
    • Hireability: HIGH — personal website explicitly states 'I am on the job market and actively seeking full-time research scientist/engineer opportunities starting in Aug 2026.'
    AG

    Animesh Garg

    medium hireability

    Assistant Professor@Georgia Institute of Technology

    Previously: Chief Science Officer @ Apptronik

    Atlanta, US

    • Stephen Fleming Early Career Professor at Georgia Tech (Atlanta, US); h-index 63; research in 3D vision, generative models, robot simulation/synthetic data (Orbit, Isaac Lab)
    • Was CSO at Apptronik (2024-2025); that role just ended and he recently updated 'Current Applications' to include Manufacturing
    • Adjacent to CAD — deep sim/synthetic-data expertise for 3D embodied AI, but not CAD-specific
    • Hireability: MEDIUM-HIGH — CSO role at Apptronik ended (2024-2025), new career motion signals visible (position_update, cv_update May 2025), manufacturing pivot suggests openness to CAD-adjacent work; however very senior and may seek leadership rather than IC researcher role
    CL

    Chao Liu

    medium hireability

    Researcher@NVIDIA

    Previously: Research And Development Scientist @ Elephas

    Madison, US

    • Generative 3D modeling researcher at NVIDIA GenAir
    • Papers on compositional 3D generation (BlobGEN-3D), camera-controllable 3D-consistent video (CamCo), and 3D point cloud self-supervised learning — directly applicable to synthetic 3D data pipelines for CAD foundation models
    • CMU PhD, h_index=16
    • Location unconfirmed (NVIDIA Research, likely Bay Area)
    • Note: LinkedIn URL in DB appears to belong to a different Chao Liu (optical imaging/Elephas researcher)
    • Hireability: MEDIUM — ~5-6 years at NVIDIA flagship research, within transition window but no job-search signals detected
    CH

    Charles Herrmann

    medium hireability

    Researcher@Google

    Previously: Postdoc @ Google

    San Francisco, US

    • Co-led Kubric (scalable synthetic dataset generator, CVPR 2022, 320 citations) and AutoFlow (learning better training sets via synthetic data generation, CVPR 2021) — direct match for synthetic data pipeline role
    • Strong 3D generative background: ZeroNVS (360° view synthesis), MonST3R (3D geometry in motion), WonderWorld (3D scene gen from single image)
    • Google Researcher in SF, h-index=22
    • Hireability: MEDIUM — was at Cornell Tech through 2020, joined Google ~2021, ~4-5 years in role (within transition window) but no explicit 'open to work' signals found
    CZ

    Chenxu Zhang

    medium hireability

    Senior Research Scientist@ByteDance

    Previously: PhD student @ The University of Texas at Dallas

    San Francisco, US

    • Senior RS at ByteDance SF Intelligent Creation Lab
    • CADDreamer (CVPR 2025 Highlight) is direct CAD object generation from single-view images — strongest possible topic match for synthetic CAD data pipelines
    • Also has strong 3D generative modeling background: MagicAnimate (325 cites, diffusion-based), Magic-Boost (multi-view 3D generation), AvatarStudio (text-to-3D)
    • PhD UT Dallas 2023, based in SF
    • Hireability: MEDIUM — ~3 years at ByteDance (joined May 2023), within transition window; ByteDance US regulatory uncertainty is a positive hireability factor; no explicit job-seeking signals but position update noted on website (Mar 2025)
    CP

    Chris Paxton

    medium hireability

    AI Innovation Lead@Agility Robotics

    Previously: Senior Lead - Embodied AI @ Hello Robot

    Pittsburgh, US

    • PhD CS (Johns Hopkins, h-index 41), strong embodied AI background across Meta AI, Hello Robot, and now Agility Robotics (AI Innovation Lead, Jan 2025)
    • Directly matches JD's 'embodied AI' experience requirement; 3D perception work (CLIP-Fields, spatial semantic representations) and large-scale robot simulation interest (robocasa fork) show adjacent skills to synthetic data pipelines
    • No direct CAD-specific work, but 3D manipulation/perception expertise is relevant
    • Pittsburgh, US — passes location constraint
    • Hireability: MEDIUM — ~16 months at Agility Robotics, past settling-in period and within typical transition window; no active open-to-work signals detected on GitHub or website
    DP

    Despoina Paschalidou

    medium hireability

    Senior Research Scientist@NVIDIA

    Previously: Postdoctoral Researcher @ Stanford University

    San Francisco, US

    • Senior Research Scientist at NVIDIA Toronto AI Lab (Santa Clara, started May 2024), specializing in 3D generative scene synthesis — ATISS (autoregressive indoor scene synthesis, NeurIPS 2021), CC3D (layout-conditioned compositional 3D scene generation, ICCV 2023), and PartNeRF (generating part-aware editable 3D shapes, CVPR 2023) directly relevant to synthetic data generation for 3D/CAD foundation models
    • ETH Zurich PhD 2021 (Computer Vision), Stanford postdoc, US-based in SF Bay Area
    • Hireability: MEDIUM — ~2 years into NVIDIA role (May 2024), within typical transition window, but no explicit job-search signals detected
    DL

    Difan Liu

    medium hireability

    Research Scientist@Adobe

    Previously: PhD student @ University of Massachusetts, Amherst

    • Research Scientist at Adobe Research with deep CAD-relevant 3D work: CSGNet (CVPR 2018, 279 cites) and Neural Shape Parsers for CSG directly address constructive solid geometry — the core CAD representation; ParSeNet (2020, 214 cites) tackles parametric surface fitting for 3D point clouds; LRM (ICLR 2024 Oral, 626 cites) builds large-scale 3D foundation model capability
    • PhD UMass Amherst (Kalogerakis lab, top-tier 3D shape analysis group)
    • Adobe Research is in San Jose CA (US)
    • Contributed to Adobe Generative Fill and Image Trace 2.0
    • Hireability: MEDIUM — ~3-4 years at Adobe (within transition window), but no explicit open-to-work signals; website is in mentor mode
    FX

    Fanbo Xiang

    medium hireability

    Senior Director, Robotics Simulation@Hillbot

    Previously: Robotics Simulation Intern @ NVIDIA

    San Diego, US

    • Creator of SAPIEN (embodied AI simulation platform used for synthetic data generation by hundreds of institutes worldwide) and ManiSkill3 (GPU-parallelized robotics sim + rendering for generalizable embodied AI, 2025)
    • Strong 3D pipeline expertise — NeuManifold (neural watertight manifold reconstruction), NeuTex (neural texture mapping), MVSNeRF (3D reconstruction)
    • PhD UCSD 2024 (Hao Su lab), deep 3D/simulation domain directly matching synthetic data pipelines for CAD/world foundation models
    • San Diego, US
    • Hireability: MEDIUM — ~18 months into Senior Director / Simulation Tech Lead role at Hillbot startup; no explicit open-to-work signals but within transition window; no website or LinkedIn activity detected
    GS

    Gopal Sharma

    medium hireability

    Senior Researcher@Samsung

    Previously: Advisor @ AuraML

    San Francisco, US

    • Deep CAD/3D modeling expertise: CSGNet (neural parser for constructive solid geometry programs, 286 citations) and ParSeNet (parametric surface fitting for 3D point clouds, 213 citations) directly address synthetic data for CAD pipelines
    • PhD UMass (3D shape understanding), postdoc UBC (neural rendering), now Senior Researcher at Samsung Research America (Sunnyvale, SF Bay Area). h-index 12
    • Hireability: MEDIUM — ~1 year into current Samsung role (joined ~April 2025 per website/CV update), no open-to-work signals, but within reach given strong topical alignment
    GS

    Guoxian Song

    medium hireability

    Staff Research Scientist@ByteDance

    Previously: Senior Research Scientist @ ByteDance

    San Francisco, US

    • Staff RS at ByteDance Bay Area, H-index 18 with SIGGRAPH/CVPR/ICCV/ICLR papers
    • Strong 3D generative modeling background (PanoHead 3D head synthesis, AvatarGen animatable avatars, diffusion-based video generation in X-Portrait/X-Dyna) — 3D generative and video/image modeling qualifications apply, though primary domain is digital humans/avatars rather than CAD or synthetic data pipelines directly. 3D NeRF and diffusion expertise is transferable
    • PhD NTU 2021
    • Hireability: MEDIUM — ~5 years at ByteDance, within typical transition window, but no explicit open-to-work signals; still actively publishing (March 2026 X-Motion demo)
    HC

    Hansheng Chen

    medium hireability

    Intern@Adobe

    Previously: Research Assistant @ Northwestern University

    San Francisco, US

    • Stanford 3rd-year PhD (co-advised by Leonidas Guibas and Gordon Wetzstein), 3D generative models and diffusion/flow expert
    • Authored Img2CAD (2025) on reverse-engineering CAD models from images — directly on-target for Spectral's CAD foundation model work
    • Also: SSDNeRF (ICCV 2023, 194 cites), 3D-Adapter (2025), Gaussian Mixture Flow Matching (ICML 2025), pi-Flow (ICLR 2026). h_index=10, CVPR 2022 Best Student Paper
    • Based in SF/Stanford; active Adobe Research intern
    • Hireability: MEDIUM — 3rd-year PhD so not yet in final transition window, no explicit job search signals, but strong industry engagement via Adobe internship and could be open to the right opportunity given the direct CAD relevance
    HT

    Hao Tan

    medium hireability

    Research Scientist@Adobe

    Previously: Research Intern @ Bloomberg

    San Francisco, US

    • Adobe Research Scientist in SF with strong 3D generative modeling portfolio — lead/co-author on LRM (620 citations), Instant3D (345), GS-LRM (207), DMV3D (205), MeshLRM (99); research focus on 3D reconstruction, multi-view generation, and virtual world generation
    • PhD UNC 2021, h-index 27. 3D generative work directly maps to synthetic data pipelines for CAD/3D foundation models per JD
    • Hireability: MEDIUM — ~4 years at Adobe (August 2021), within typical transition window; no explicit open-to-work signals but no negative signals either
    HB

    Homanga Bharadhwaj

    medium hireability

    Research Scientist@Meta

    Previously: Student Researcher @ DeepMind

    Seattle, US

    • Strong embodied AI/robotics RS at Meta Reality Labs (Seattle, US)
    • Work spans synthetic data augmentation (RoboAgent: semantic augmentations to multiply datasets), video generation for robot policy learning (Gen2Act: human video generation as synthetic demonstrations), and diffusion-based imitation (DemoDiffusion)
    • No direct CAD/3D modeling experience, but embodied AI is explicitly a qualifying criterion in the JD, and their data-generation-for-learning work transfers well to synthetic data pipelines
    • H-index 23, CMU PhD
    • Hireability: MEDIUM — likely 2-3 years at Meta Reality Labs (joined post-PhD ~2023), within typical transition window; no explicit 'open to work' signals found
    JH

    Jiahui Huang

    medium hireability

    Senior Research Scientist@NVIDIA

    Previously: Research Scientist @ NVIDIA

    San Francisco, US

    • Senior RS at NVIDIA SF with strong 3D generative modeling background directly relevant to synthetic data for CAD foundation models — XCube (large-scale 3D generation with sparse voxel hierarchies, 2024), GEN3C (3D-informed world-consistent video generation, CVPR 2025 Highlight), TerraCraft (city-scale generative procedural modeling with NL, 2025). h_index 21, PhD Tsinghua CS 2023
    • Located in San Francisco
    • Hireability: MEDIUM — ~3 years at NVIDIA (RS since 2023, promoted to Senior RS ~2025), recently promoted which suggests settling in; no open-to-work signals; website updated Dec 2025 with paper additions only; actively recruiting interns
    JT

    Jonathan Tremblay

    medium hireability

    Researcher@NVIDIA

    Previously: PhD student @ McGill University

    • Pioneer of synthetic data for robotics at NVIDIA — created NDDS (NVIDIA Deep Learning Dataset Synthesizer), co-authored domain randomization paper (CVPR 2018, 1,248 citations), FactorSim generative simulation (NeurIPS 2024), GRS (generating robotic simulation tasks from real-world images, CVPR 2025), and 3D-GENERALIST (VLA for crafting 3D worlds, 3DV 2026). h_index 35, PhD McGill CS
    • Research focuses precisely on synthetic data pipelines for training CV/robotics models — exact match for JD
    • Location unconfirmed (NVIDIA @nvidia.com, most co-authors at NVIDIA Seattle, no explicit US address found)
    • Hireability: MEDIUM — ~7+ years at NVIDIA with no active job-seeking signals found; NVIDIA restructuring provides some upside
    KL

    Kejie Li

    medium hireability

    Research Engineer@Meta

    Previously: Research Scientist @ ByteDance

    Redmond, US

    • 3D computer vision expert at Meta Reality Labs (Redmond WA, US) with strong generative 3D background — co-authored MVDream (multi-view diffusion for 3D generation) and MVLight (text-to-3D generation), directly relevant to JD's '3D domains/generative modeling' requirement
    • PhD from Adelaide, postdoc at Oxford, h_index 17
    • No direct CAD or synthetic data pipeline experience visible
    • Hireability: MEDIUM — Career history shows 1.5-2yr job switches (Oxford 2yr → ByteDance 1.5yr → Meta now 1.5yr), currently 1.5yr at Meta with no explicit open-to-work signal, but pattern suggests possible transition window
    LS

    Liangchen Song

    medium hireability

    Engineer@Apple

    Previously: Intern @ InnoPeak Technology

    San Francisco, US

    • Strong 3D generative researcher at Apple SF — published on text-driven 3D scene synthesis (RoomDreamer, 2023), dynamic NeRF (NeRFPlayer, TVCG'23, 304 citations), single-image novel-view synthesis (Efficient-3Dim, 2024), and scalable video generation (STIV, Apple, 2024)
    • Has prior synthetic image experience
    • No explicit CAD work but directly relevant 3D generative pipeline expertise (h_index=19)
    • Hireability: MEDIUM — ~2-3 years at Apple (joined post-PhD ~2022-2023), within typical transition window; no explicit job-seeking signals found
    LM

    Long Mai

    medium hireability

    Senior Research Scientist@Adobe

    Previously: Senior Research Scientist @ ByteDance

    San Francisco, US

    • Senior RS at Adobe Research SF (2023-present), specializing in 3D generative modeling — MVDream co-author (multi-view diffusion for 3D object generation, ICLR 2024) and BlockGAN (3D object-aware scene representations)
    • Active output at ICCV/SIGGRAPH 2025 (video tokenizers, controllable video generation)
    • PhD CS Portland State 2017, h-index 26
    • Strong fit for synthetic data pipelines powering 3D/CAD models — no direct CAD experience but multi-view 3D generation is the core technical skill
    • Hireability: MEDIUM — ~2.5 years at Adobe (second stint, previously at ByteDance/SEA AI Lab), within typical transition window; no explicit job-seeking signals from pipeline or website
    SB

    Sai Bi

    medium hireability

    Senior Research Scientist@Adobe

    Previously: Research Intern @ Meta

    San Francisco, US

    • PhD (UCSD/Ramamoorthi) + Senior RS at Adobe SF
    • Two papers directly on synthetic data for 3D: 'LRM-Zero: Training Large Reconstruction Models with Synthesized Data' (2024) and 'MegaSynth: Scaling Up 3D Scene Reconstruction with Synthesized Data' (2025) — exactly the skill set Spectral needs
    • Prolific 3D foundation model work: LRM (620 citations), Instant3D, GS-LRM, DMV3D, OpenRooms synthetic dataset framework. h_index 28
    • Hireability: MEDIUM — ~4-5 years at Adobe post-PhD (2021), in transition window, but no open-to-work signals visible; website updates only new papers, no career motion
    SY

    Siwei Yang

    medium hireability

    Ph.D. student@UC Santa Cruz

    Previously: Research summer intern @ CCVL @JHU

    San Francisco, US

    • PhD student at UC Santa Cruz specializing in synthetic dataset generation with generative models — directly relevant to the role
    • Key work: HQ-Edit (94 citations, instruction-based image editing dataset) and GPT-image-edit-1.5M (GPT-generated million-scale image dataset), demonstrating hands-on synthetic data pipeline experience
    • No 3D/CAD background found; work is 2D image-focused
    • Based in SF, US
    • Hireability: MEDIUM — active recent GitHub commits (April 2026), no explicit availability signals, PhD year unknown but research output suggests mid-to-late stage
    TG

    Thibault Groueix

    medium hireability

    Research Scientist@Meta

    Previously: Research Scientist @ Adobe

    San Francisco, US

    • AI Research Scientist at Meta SF, focused on 'training large-scale models for 3D generative AI, towards building world models.' Foundational 3D generative work: AtlasNet (CVPR 2018, 1678 citations), TextDeformer (text-guided geometry manipulation), CNOS (CAD-based object segmentation), MagicClay (generative mesh sculpting), and SuperFrusta (residual primitive fitting for 3D shapes, CVPR 2026 oral) — directly relevant to CAD foundation models. h-index 16, PhD from École des Ponts
    • Location is SF, hard constraint met
    • Hireability: MEDIUM — established RS at Meta, very actively publishing (CVPR 2026, SIGGRAPH Asia 2025), no pipeline signals of job-seeking, but ~2-3 years into Meta role puts him within the typical transition window
    TW

    Tingwu Wang

    medium hireability

    Senior Research Scientist@Nvidia

    Previously: Senior Research Scientist @ Nvidia

    San Francisco, US

    • Staff RS at NVIDIA GEAR working on generative embodied AI — humanoid motion generation (MotionBricks SIGGRAPH'26, SONIC/CHIP arXiv 2025, Kimodo), physics simulation, and physics-based character-scene synthesis
    • Directly hits the JD's 'embodied AI / world modeling' criterion
    • PhD CS UofT 2022, h_index 11, SF-based per LinkedIn (GitHub shows Toronto — possible discrepancy)
    • Hireability: MEDIUM — ~4 years at NVIDIA (within transition window), CV updated March 2026 with 2 position_updates in last 6 months suggesting career motion, but no explicit open-to-work signal
    AH

    Ankur Handa

    low hireability

    Principal Scientist@NVIDIA

    Previously: Research Scientist @ Imperial College London

    • Principal Research Scientist at NVIDIA Robotics co-leading Dexterity team; created SceneNet (5M synthetic indoor images, CVPR 2016 + ICCV 2017) and Isaac Gym (GPU physics sim for robot learning, NeurIPS 2021) — landmark synthetic data / simulation work directly relevant to Spectral's pipeline
    • Has personal `claude-cad` GitHub repo (using Claude + build123d for CAD generation). h-index 35, PhD Imperial College London
    • Location likely US (previously at OpenAI SF; GitHub 'London' / 'University of Cambridge' appears stale/outdated)
    • Hireability: LOW — 4+ years at NVIDIA co-leading a named research team, no transition signals (no LinkedIn history, website updated Feb 2026 for cosmetic tweaks only, pipeline shows no activity)
    AM

    Arsalan Mousavian

    low hireability

    Research Scientist@George Mason University

    Previously: Robotics Research Manager @ NVIDIA

    Seattle, US

    • Ex-NVIDIA Robotics Research Manager, now 'Building Physical AI'; co-authored Cosmos World Foundation Model (2025) and PointWorld 3D world model (Jan 2026)
    • Has dedicated synthetic data publication ('Synthesizing Training Data for Object Detection', 2017), built ACRONYM large-scale simulation dataset (2021), and strong 3D/embodied AI background (6-DOF GraspNet, pose estimation, object manipulation). h_index 38, senior-level researcher
    • Located in Seattle, US (passes US constraint, not SF)
    • Hireability: LOW — Jan 2026 paper still shows NVIDIA affiliation; no open-to-work signals; may still be at NVIDIA on Physical AI team or recently departed to build a startup; pre-computed hireability low with low confidence
    CH

    Chen Huang

    low hireability

    Senior Deep Learning Research Engineer@Apple

    Previously: Deep Learning Research Engineer @ Apple

    San Francisco, US

    • Senior DL Research Engineer (Research Scientist) at Apple SF, explicitly focused on 'Video & 3D Generative AI'
    • Recent ICML 2025 paper (Cavia) on camera-controllable multi-view video diffusion directly applicable to 3D synthetic data generation for CAD models
    • PhD ECE, h_index 24
    • Hireability: LOW — ~9 years at Apple with no pipeline signals of career motion (no LinkedIn changes, no website activity detected); entrenched in current role
    DF

    David Fouhey

    low hireability

    Assistant Professor@New York University

    Previously: Assistant Professor @ University of Michigan

    New York, US

    • Strong 3D CV researcher at NYU — 3D reconstruction from images, surface normal estimation, PixelSynth (generative 3D), hand/object interaction; h-index 35
    • Also affiliated with Polymathic AI (foundation models)
    • NY-based, meets US constraint
    • Hireability: LOW — Institute Associate Professor at NYU (settled academic position), no pipeline signals or career motion, website actively not recruiting students/postdocs
    FL

    Fangchen Liu

    low hireability

    Research Scientist@DeepMind

    Previously: Research Intern @ Google

    San Francisco, US

    • Strong 3D + world model + embodied AI RS at Google DeepMind SF
    • PhD Berkeley (Abbeel lab, h=20)
    • Published SAPIEN (675 cites, simulated 3D part-based environment), Masked World Models, and 'Evaluating Gemini Robotics Policies in a Veo World Simulator' — directly relevant to synthetic data pipelines for 3D/CAD foundation models
    • Hireability: VERY LOW — website updated 2026-04-21 (8 days ago) announcing transition from final-year PhD to RS at DeepMind; just started
    FL

    Fujun Luan

    low hireability

    Staff Machine Learning Researcher@Apple

    Previously: Research Scientist 2 @ Adobe

    San Francisco, US

    • Staff ML Researcher at Apple (SF, 2025–present) with strong 3D generative background directly relevant to synthetic data pipelines for CAD: authored MeshLRM (high-quality mesh reconstruction, 2025), DMV3D (multi-view diffusion for 3D, 2024, 205 citations), PF-LRM (pose-free shape prediction, 2024), and LVSM (large view synthesis, 2025)
    • PhD Cornell 2021, h_index=26
    • Prior 3+ years at Adobe doing neural rendering and 3D
    • Hireability: LOW — ~1 year into Apple role, just promoted to Staff level (pipeline signals show title change Jan 2026); current headline emphasizes 'Post-training & Foundation Models' indicating settled into a new Apple project; no open-to-work signals
    GP

    Georgios Pavlakos

    low hireability

    Assistant Professor@University of Texas at Austin

    Previously: Postdoctoral Researcher @ University of California, Berkeley

    Austin, US

    • Strong 3D vision researcher with directly relevant work: MegaSynth (2024, synthetic data for 3D scene reconstruction), Atlas Gaussians Diffusion for 3D generation (2025), and InstantSplat Gaussian splatting (2024). h_index=31, PhD from UPenn, postdoc at UC Berkeley
    • Austin TX (US, not SF)
    • Hireability: LOW — Assistant Professor at UT Austin since Jan 2024, ~2.5 years into tenure-track position, no pipeline signals of industry interest or career motion
    HJ

    Haian Jin

    low hireability

    Student Researcher@DeepMind

    Previously: Research Intern @ Adobe

    San Francisco, US

    • Direct hit on synthetic data for 3D: MegaSynth (2024) scales 3D scene reconstruction with synthesized data
    • Strong 3D generative modeling portfolio: One-2-3-45 (NeurIPS 2023, 612 citations), LVSM (ICLR 2025 Oral, large view synthesis), Neural Gaffer (NeurIPS 2024 relighting via diffusion), TensoIR (CVPR 2023)
    • Cornell PhD advised by Noah Snavely, Google PhD Fellowship 2025, currently Student Researcher at Google DeepMind
    • US-based (Cornell/NY area)
    • Hireability: LOW — only 2-3 years into a 5-year PhD (expected 2028), prestigious Google Fellowship recipient, embedded at DeepMind; unlikely to leave PhD program early
    HJ

    Hanwen Jiang

    low hireability

    Research Scientist@Adobe

    Previously: Research Intern @ Adobe

    Austin, US

    • PhD UT Austin 2025, now Research Scientist at Adobe Research (Austin, TX)
    • Led MegaSynth (CVPR 2025) — synthetic data pipeline for scaling 3D scene reconstruction — which directly matches the synthetic data + 3D domain focus
    • Also published Real3D (ICCV 2025) and RayZer (ICCV 2025, Best Student Paper HM) on scalable 3D reconstruction and world modeling
    • Strong fit for synthetic data + 3D generative pipeline work
    • US-based (Austin TX)
    • Hireability: LOW — started at Adobe Research ~June 2025 (~10 months ago), very fresh hire; no open-to-work signals on GitHub, LinkedIn, or website
    JL

    Jiahao Li

    low hireability

    PhD student@TTIC

    Previously: Intern @ Adobe

    • PhD from TTIC (Greg Shakhnarovich) focused on 3D generation and structure from motion — directly on target for synthetic data pipelines for CAD/3D foundation models
    • Published Instant3D (ICLR 2024, fast text-to-3D), Score Jacobian Chaining (CVPR 2023, diffusion-based 3D), and FastMap (3DV 2026 Best Paper Candidate)
    • Now Senior ML Engineer at Tesla AI in Bay Area (US, preferred geo)
    • Hireability: LOW — started at Tesla AI ~6 weeks ago (March 2026 per website commit history), very recently settled in
    JS

    Joshua M. Susskind

    low hireability

    Distinguished Engineer@Apple

    Previously: Research Manager, Deep Learning Scientist @ Apple

    San Francisco, US

    • Apple SF Distinguished Engineer with exactly on-point expertise for this role: Hypersim (525-citation photorealistic synthetic dataset for 3D scene understanding, 2021), '3D Shape Tokenization via Latent Flow Matching' (2025), GAUDI/Control3diff (3D generative scene models), and explicit 'world models' research expertise
    • Research directly spans synthetic data pipelines, 3D shape generative modeling, and world modeling — the trifecta this role needs
    • PhD Toronto 2009, h_index 39
    • SF-based
    • Hireability: LOW — Apple Distinguished Engineer is a deeply senior, high-compensation IC role with no open-to-work signals, no LinkedIn/website changes in pipeline, ~16 yrs industry tenure; would be a steep hill to recruit
    KS

    Kalyan Sunkavalli

    low hireability

    Principal Scientist@Adobe

    Previously: Senior Research Manager @ Adobe

    San Francisco, US

    • Principal Scientist at Adobe Research SF (h_index=60, PhD Harvard CS) with deep 3D generative modeling and synthetic data expertise: co-authored OpenRooms (CVPR 2021 oral), a framework for photorealistic synthetic indoor scene datasets; led LRM, Instant3D, GS-LRM, and DMV3D — 3D reconstruction/generation models directly applicable to CAD foundation model pre-training
    • Research expertise in 3D vision, neural rendering, and procedural material generation aligns strongly with Spectral's synthetic data pipeline needs
    • Hireability: LOW — ~13+ years at Adobe Research, actively recruiting interns there, no open-to-work signals from pipeline signals or personal website
    KG

    Kyle Genova

    low hireability

    Senior Research Scientist@DeepMind

    Previously: Senior Research Scientist @ Google

    New York, US

    • Staff RS at Google DeepMind (Funkhouser's team, Foundational Research Unit), PhD Princeton 2021 in 3D deep learning
    • Core 3D shape/scene modeling expertise (LDIF, structured implicit functions, CvxNet learnable convex decomposition) directly applicable to 3D synthetic data and CAD foundation models
    • Currently working on Gemini pretraining for 3D understanding; ICCV 2025 papers on 3D VQA and scene motion generation. h-index 21, multiple CVPR/ICCV orals
    • NY-based, passes US constraint
    • Hireability: LOW — recently promoted to Staff RS (Nov 2025 per LinkedIn), no open-to-work signals, working on flagship Gemini pretraining at DeepMind
    MH

    Milos Hasan

    low hireability

    Research Scientist@NVIDIA

    Previously: Researcher @ Adobe

    • Principal Scientist at NVIDIA Research (Real-Time Graphics Group, joined Jul 2025), PhD Cornell, h-index 34
    • Strong 3D generative background directly relevant to synthetic data for 3D: OpenRooms (photorealistic synthetic indoor scene dataset framework, CVPR 2021), MaterialGAN + MatFormer (generative 3D material models), RGB<>X (diffusion models for material/scene synthesis, 2024)
    • Expertise in inverse rendering and neural materials maps well to synthetic data pipelines for 3D foundation models — fits 'other 3D domains' qualification
    • Not CAD-specific but adjacent
    • Location: NVIDIA Research HQ Santa Clara (Bay Area), satisfies US requirement
    • Hireability: LOW — only ~9 months into new NVIDIA Research role as of Apr 2026, unlikely to be exploring options so soon
    PW

    Peng Wang

    low hireability

    Senior Staff Research Scientist (Tech Lead)@ByteDance

    Previously: Staff Research Scientist (Tech Lead) @ ByteDance

    San Francisco, US

    • MVDream, MVDiffusion, and MVLight co-author — multi-view diffusion models for 3D generation (2023-2024), h-index 51, PhD UCLA
    • Senior Staff RS (Tech Lead) at ByteDance AI Lab in Sunnyvale/SF
    • Strong 3D generation and multi-view synthesis background directly applicable to synthetic data pipelines for CAD/3D foundation models
    • Hireability: LOW — 6+ years at ByteDance with no pipeline signals, no LinkedIn changes, and no website/GitHub activity suggesting mobility
    SL

    Sifei Liu

    low hireability

    Senior Research Scientist@NVIDIA

    Previously: Research Scientist @ NVIDIA

    San Francisco, US

    • Principal Research Scientist & Tech Lead at NVIDIA Research SF (LPR team, h-index 44)
    • Strong 3D generative portfolio: COLMAP-Free 3DGS (CVPR 2024, 232 citations), AGG (Amortized 3D Gaussians, 2024), No Pose No Problem 3DGS (ICLR 2025)
    • Deeply involved in Cosmos and Isaac GR00T VLM/VLA foundation models
    • Directly hits JD's '3D domains + embodied AI + world modeling' requirements; no direct CAD/synthetic-data pipeline work but strong adjacent 3D generative skillset
    • Hireability: LOW — just promoted to Principal Research Scientist / Tech Lead on 2026-03-23, deeply embedded in NVIDIA flagship products (Cosmos, GR00T); no market signals detected
    YH

    Yicong Hong

    low hireability

    Research Scientist@Adobe

    Previously: Researcher @ Adobe

    San Francisco, US

    • RS at Adobe Research (2024–2025) with strong 3D generative modeling background — co-authored LRM, Instant3D, and Long-LRM (Gaussian splats); also published on scaling synthetic data generation for embodied navigation
    • PhD ANU 2023, h-index 20, Bay Area
    • Directly matches JD's 3D domain + video/world modeling + synthetic data pipeline goals
    • Hireability: LOW — recently left Adobe and joined a stealth startup (GitHub company field + website both updated; multiple position_update events in H2 2025, most recent ~Feb 2026); likely 4-5 months into new role

    Runs

    #1completed0 qualified / 0 foundApr 29, 8:36 PM