Founding Research Engineer Research · · San Francisco, CA $160K - $250K 1-2%

Adam; San Francisco (CA); Hybrid; USD 160,000 - 250,000; Full time

Job summary

Adamcad is assembling the founding research team to advance AI for parametric CAD in 3D space. The role focuses on training models that interpret natural language to generate CAD, and on building robust representations and evaluation methods for geometric accuracy.

You will own the full AI stack from data collection to training, evaluation, and inference. We seek contributors with strong ML credentials who can ship systems and push the boundaries of LLMs in CAD applications, reporting to the

Qualifications

• Must have hands-on experience building ML models for 3D or CAD tasks.
• Ability to own end-to-end ML stack from data collection to inference.
• Proven track record of original contributions in ML (papers, open source, or shipped systems).

Responsibilities

• Train models that generate CAD from natural language.
• Build effective vector representations of CAD features.
• Design evals to measure geometric accuracy in 3D space.
• Optimize inference for faster CAD edits.
• Create RL environments to teach long-horizon agents to perform design tasks end-to-end.

Skills

AWS

Python

PyTorch

Machine Learning

Deep Learning

Job description

San Francisco, CA, US Full-time US citizen/visa only 3+ years Amazon Web Services (AWS) Python Torch/PyTorch Machine Learning Deep Learning

Adam is an applied research and product lab building the future of computer-aided design, where engineers will be able to speak physical objects into existence. With Adam, you’ll use AI to complete design tasks end-to-end. For example, select a face and say, “Add mounting holes matching the bolt pattern from the other part, with identical diameters, spacing, and offsets,” and Adam handles the geometry, selects standard hardware, and outputs production-ready CAD.

We just wrapped up YC’s W25 batch, built a state-of-the-art text-to-CAD interface, and had one of the most viral YC launches of all time.

Our first product made CAD approachable and fun with AI, bringing tens of thousands of people into 3D printing. Now we’re enhancing the capabilities of engineers to create feature-rich models compatible with industry-standard software.

We’re a small technical team of engineers, designers, and researchers who love CAD and building things. Our founders are UC Berkeley alumni, second-time founders, and early employees at major AI startups.

We're in-person at the Adam House in the SF Marina, powered by green tea, lots of natural light, and a wall of 3D printers.

If you’re excited about shipping fast and shaping the future of how things are made, come join us.

About the role

We’re building the founding research team at Adam.

At Adam, we’re tackling a frontier problem: training AI models to intelligently interpret and edit parametric CAD in 3D space. This demands creativity, deep technical ability, and novel thinking.

As part of the founding research team, you may:

• Train models that generate CAD from natural language
• Build effective vector representations of CAD features
• Design evals to measure geometric accuracy in 3D space
• Optimize inference for faster CAD edits
• Create RL environments to teach long-horizon agents to perform design tasks end-to-end

We are looking for people who:

• Are excited to explore and push the boundaries of LLMs, especially in the domain of 3D and CAD generation
• Have trained and served state-of-the‑art models.
• Can own the full AI stack, from data collection to model training, evaluation, and inference
• Have a track record of original contributions in ML (first-author papers, open‑source projects, or shipped systems)
• Operate with high agency and own problems end‑to‑end

At Adam, you will be given the autonomy to do some of the best work of your life and redefine how the physical world is built.

San Francisco, CA, US Full-time $200K - $250K 1.00% - 2.00% 3+ years

San Francisco, CA, US Full-time $160K - $250K 1.00% - 2.00% 1+ years