Research Engineer, Synthetic Data

HUD; San Francisco (CA); On-site; USD 140,000 - 210,000; Full time

Benefits offered by this job

Medical + dental + vision

Lunch and dinner in office

Holiday break

Equinox membership

401k

Commuter benefits

AI tokens access

Job summary

HUD is seeking Research Engineers to build our synthetic data pipeline. You’ll turn domain‑specific workflows into synthetic training tasks that are realistic enough to teach useful behavior, structured enough to generate at scale, and difficult enough to expand model capabilities.

You will work with subject‑matter experts to create synthetic tasks for training AI agents across a range of professional and technical domains, design generation methods, and build tooling to mutate, validate, and

Qualifications

• Proficiency in Python, Docker, and Linux environments.
• Experience with synthetic data research methods – please elaborate in your application.
• Strong understanding of what “good synthetic data” means and its limitations.
• Built synthetic data pipelines end‑to‑end without a fully prescribed roadmap.

Responsibilities

• Work with subject‑matter experts to create synthetic tasks for training AI agents across a range of professional and technical domains
• Design synthetic task generation methods that produce diverse, realistic, and learnable tasks
• Build systems and tooling to mutate, validate, and improve synthetic tasks
• Analyze model and agent performance on synthetic tasks to understand what the tasks are teaching and where they fail
• Develop metrics to quantify and understand synthetic task diversity, realism, learnability, etc.

Skills

Python

Docker

Linux

Synthetic data

Job description

About the role

We’re looking for Research Engineers to build our synthetic data pipeline. You’ll turn domain‑specific workflows into synthetic training tasks that are realistic enough to teach useful behavior, structured enough to generate at scale, and difficult enough to expand model capabilities.

Responsibilities

• Work with subject‑matter experts to create synthetic tasks for training AI agents across a range of professional and technical domains
• Design synthetic task generation methods that produce diverse, realistic, and learnable tasks
• Build systems and tooling to mutate, validate, and improve synthetic tasks
• Analyze model and agent performance on synthetic tasks to understand what the tasks are teaching and where they fail
• Develop metrics to quantify and understand synthetic task diversity, realism, learnability, etc.

Experience

You may be a good fit if you have:

• Proficiency in Python, Docker, and Linux environments
• Experience with synthetic data research methods – please elaborate in your application
• Strong understanding of what “good synthetic data” means and its limitations
• Built synthetic data pipelines end‑to‑end without a fully prescribed roadmap
• Experience working on environments, evals, and benchmarks

Strong candidates may also:

• Be detail‑oriented and able to spot subtle inconsistencies or edge cases in synthetic data
• Be able to reason from first principles about task design, scoring, and failure modes
• Thrive in unstructured problem spaces
• Have early‑stage startup experience with ability to work independently in fast‑paced environments
• Strong communication skills for remote collaboration across time zones

Logistics

• Employment: Full‑time.
• Location: On‑site only, for now. You can join the team in the San Francisco Bay Area or Singapore offices.
• Visa Sponsorship: We provide support for relocation and visas for strong full‑time candidates to the US or Singapore.
• Timeline: Applications are rolling. The process is 2 technical interviews and a 2‑3 day work trial.

What we offer

• Competitive compensation
• 100% covered top‑of‑the‑line medical, dental, and vision from Blue Shield of CA (US employees)
• Lunch and dinner when you’re in the office
• Company‑wide holiday break (Christmas Eve to New Year’s Day) on top of PTO and paid holidays
• Other perks including an Equinox membership, 401k, and commuter benefits (US employees)
• Unlimited access to tokens for ChatGPT, Claude Code, Cursor, etc. (by unlimited, we mean no one on our token usage leaderboard has ever hit a limit. So we have no idea what the limit is.)