READ THIS BEFORE YOU APPLY

Please open the Tracer GitHub repo (https://github.com/adrida/tracer), read the README, and leave a star if the project resonates with you. In your application, include one idea, question, or observation you had while exploring the repo. No need to open PRs or build anything. A thoughtful sentence is enough.

---

Hiring: Founding ML / AI Engineering Interns at Tracer

Tracer is a San Francisco-based AI startup building open-source tools that help AI teams cut LLM inference costs. We train a lightweight ML model on your own production LLM traces and handle the repeated calls for near-zero cost. Same prompts, same providers, up to 95% lower inference bill.

We are looking for strong builders interested in machine learning, LLM applications, Python, backend engineering, embeddings, evaluation, developer tools, and open-source software.

This is a founding internship role. You will work directly on the core product, contribute to the open-source repo, and help improve the systems behind Tracer's surrogate training, parity gates, routing logic, benchmarks, and developer experience.

Responsibilities

- Build and improve Python pipelines for ML surrogate training, routing, clustering, embeddings, vector search, and evaluation

- Work on LLM routing logic, parity gates, and quality monitoring

- Improve the developer experience around the open-source SDK, documentation, examples, and integrations

- Explore benchmarks, datasets, and evaluation frameworks for LLM classification and routing

- Contribute ideas, identify gaps, and help shape the product from an early stage

Who we are looking for

- Students, recent graduates, self-taught engineers, researchers, or early builders

- Strong interest in ML, LLMs, AI infrastructure, developer tools, or open-source

- Comfortable reading unfamiliar code, learning quickly, and working independently

- Experience with Python is strongly preferred

- Experience with ML projects, backend engineering, embeddings, vector search, or open-source contributions is a plus

Role details

- Internship, paid

- Remote worldwide

- 3-6 months, flexible start date

- Work directly with the founder

How to stand out

Open the Tracer GitHub repo, read the README, and leave a star if the project resonates with you. In your application, include one idea, question, or observation you had while exploring it. No PRs, no issues, no free work expected.

GitHub: https://github.com/adrida/tracer

Apply: https://tracerml.ai/hiring/

Pay: ₹393,473.71 - ₹1,630,962.62 per year

Benefits:

• Flexible schedule
• Work from home

Work Location: Remote