> Markdown version of [/jobs/ext/4625167-machine-learning-engineer](https://www.wearedevelopers.com/jobs/ext/4625167-machine-learning-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Machine Learning Engineer

- **Company:** Clair Health, Inc.
- **Location:** Menlo Park, CA, United States
- **Experience:** Experienced
- **Salary:** $130,000 - $170,000
- **Contract:** Permanent contract
- **Skills:** Firmware, Python (Programming Language), Machine Learning, Open Source Technology, Signal Processing, Large Language Models, Deep Learning, Model Validation, Machine Learning Operations, Virtual Agents
- **Published:** November 7, 2026
- **Apply:** https://www.juju.com/job/28_25303_291bf390a3ec3987dcfd8e40db40be6c

## About the Role

* 2 to 5 years building ML systems that ran in production or on real users' data.
 * Strong Python, a modern deep learning framework, and the scientific Python stack.
 * Experience with time-series or physiological data, such as PPG, ECG, IMU or EEG, or with LLM applications. Ideally both, or a clear desire to learn the side you know less.
 * Careful about evaluation. You split by subject, look for leakage, check calibration and test against simple baselines.
 * You want to work in person with a small team in Menlo Park., * A background in signal processing, biomedical engineering or a related research field.
 * Experience with wearables, sleep or cardiovascular physiology.
 * On-device ML: model compression and mobile deployment.
 * Experience with agents, retrieval, or real-time voice systems.
 * Clinical study or medical device experience.
 * Publications, or open-source work you can show us.
 * Interest in women's health.

## Description

This is a role for someone who wants to work on the whole path from raw sensor data to something a woman can act on. Clair's band records biosignals day and night. As our Machine Learning Engineer, you will work on the models that turn those signals into heart, sleep, activity and hormone insights, and the AI agent that helps each member understand them. As our Machine Learning Engineer, you will join the team that built them and own parts of that system: finding where models break on real-world data, validating them on study data, and shipping better versions. You will work directly with the founders, and your work will be what members see at launch., This is a role for someone who wants to work on the whole path from raw sensor data to something a woman can act on. Clair's band records biosignals day and night. As our Machine Learning Engineer, you will work on the models that turn those signals into heart, sleep, activity and hormone insights, and the AI agent that helps each member understand them. As our Machine Learning Engineer, you will join the team that built them and own parts of that system: finding where models break on real-world data, validating them on study data, and shipping better versions. You will work directly with the founders, and your work will be what members see at launch.

What you'll do

Biosignals and sensor models

 * Own and improve our existing biosignal models on raw signals: heart and sleep metrics, activity, signal quality and cycle-related physiology.
 * Find where they fail on real-world wearable data, such as motion artifacts, poor contact, gaps, unusual wear patterns and hardware or firmware changes, and fix the root cause.
 * Validate models on data from our studies and against reference devices with subject-level evaluation, and retrain them as labelled data grows.
 * Build new models where the product needs them, in the cloud or on device.

Agents, voice and LLMs

 * Improve Clair's AI agent, which answers members' questions grounded in their own wearable and cycle data.
 * Improve voice and conversational experiences in the app.
 * Extend our evals and guardrails for a health product: grounding, factual accuracy, safety around fertility and medical topics, latency and cost.

Research to production

 * Read papers and new methods, test them honestly against what we already have, and adopt what adds value.
 * Write production Python, and work with our MLOps and fullstack engineers to ship, monitor and improve your models.
 * Contribute to study design and analysis as we collect labelled data for training and validation.