Founding Engineer (Full Stack, ML DevTools & Systems)

HUD•San Francisco, CA

3d•$150,000 - $240,000

About The Position

At HUD, we’re building the future of how companies and individuals train and evaluate AI. We believe that in the near future, most post-training data used to align and improve LLMs will flow through HUD. We build a platform and developer tools that let teams create post-training data through RL environments and run reinforcement fine-tuning (RFT) reliably, reproducibly, and at scale. We’re trusted by foundation labs, Fortune 500s, and fast-growing startups. We’re also a high-caliber team: former founders, published ML researchers, Olympiad medalists, and engineers who have built products with real adoption. We run lean, move fast, and hold an extremely high bar. We’re hiring a Founding Engineer to be a core contributor across our platform, SDK, and developer experience. This is a high-ownership role at the center of the product: you’ll build primitives, APIs, and workflows that ML engineers and researchers use daily to create data, run post-training, and evaluate systems. You’ll operate across the stack—from Python SDK design to backend systems and infrastructure—while staying relentlessly product-minded. You’ll also work directly with customers (engineers and researchers at startups, enterprises, and frontier labs) to understand their workflows, unblock them, and ship what they actually need. If you’ve ever cared deeply about how ML tools feel to use—ergonomics, reliability, abstractions, and “it just works” workflows—this role is for you.

Requirements

Strong production experience in Python.
Comfort across the stack (APIs, data systems, frontend integration where needed).
Deep understanding of Docker and Linux environments; strong debugging ability.
Cloud competence (K8s and AWS fundamentals—compute, networking, storage, IAM).
Strong product instincts and a bias toward shipping.
Ability to write clean, maintainable code with strong taste in interfaces and DX.

Nice To Haves

Experience with reinforcement learning, post-training, or model training workflows.
Experience building or using LLM/agent eval frameworks (Inspect, EleutherAI tooling, custom harnesses).
Experience designing SDKs, CLIs, or developer platforms.
Kubernetes experience (deployment, scaling, job orchestration).
Active participation in the ML community (open-source contributions, writing, research engagement, etc.).

Responsibilities

Build the core HUD platform
Design and implement backend systems for post-training workflows (RFT jobs, dataset/data flow primitives, run tracking, artifacts, permissions).
Build reliable execution and orchestration primitives with strong isolation and reproducibility.
Own the SDK and developer experience
Build and iterate on our Python SDK: clean APIs, excellent docs, great errors, sharp defaults, and extensibility.
Create “golden path” workflows for common user goals: creating post-training data, launching RFT runs, evaluating results, and iterating quickly.
Ship eval-native workflows
Help build eval pipelines for LLMs/agents that connect naturally to post-training loops (capability measurement → data creation → training → re-eval).
Go deep on systems and reliability
Build with Docker, Linux, and cloud infrastructure in mind; ensure consistent environments across local, CI, and production.
Improve performance, observability, and debuggability of job execution and data pipelines.
Contribute to Kubernetes deployment patterns and scaling.
Work directly with customers
Partner with engineers and researchers at startups, enterprises, and foundation labs.
Turn messy real-world feedback into product improvements: better abstractions, missing primitives, clearer docs, smoother onboarding.