Latest AI JobsBrowse Jobs
M

Research Engineer, Post-training

medraaiSan FranciscoEngineering
Apply on company site
Salary
Not listed
Job type
Full-time
Posted
1 Oct 2026
Status
Open · checked 7d ago

About this role

Research Engineer, Post-training at medraai in San Francisco is an early-stage role for an engineer with a strong foundation in machine learning and practical system deployment. This position suits an experienced technologist eager to develop AI models for automated scientific laboratories and biopharma partnerships.

Role Overview

medraai is hiring a Research Engineer, Post-training in San Francisco to create Physical AI Scientists and robotic systems for the life sciences industry. The organization recently secured a $52M Series A funding round and operates an autonomous lab facility in the United States.

Key Responsibilities

Design post-training methodologies for AI models by selecting problem focuses, designing data collections, executing machine learning experiments, and deploying models into production.
Construct and manage data pipelines that combine internal sources and public data for post-training.
Design evaluation frameworks to measure model performance on scientific protocols and assay development.
Build agentic systems featuring custom tool calls and context management to generate scientific insights in laboratory settings.
Collaborate with robotics engineers, operations staff, and scientists to integrate reasoning capabilities into live experimental loops.
Influence the technical direction and engineering culture of the machine learning team.

Requirements

Practical background in deploying AI-driven workflows in real-world environments.
Problem-solving capabilities focused on debugging complex systems.
Understanding of probability, statistics, and foundational machine learning concepts.
Competence in managing the entire post-training stack, including RL environments, data pipelines, and agentic evaluations.
Programming proficiency in Python along with experience using at least one deep learning framework such as PyTorch or JAX.
Background working with large language models, reinforcement learning, post-training, or agentic architectures.

Skills and Tools

Python
PyTorch
JAX
Large Language Models (LLMs)
Reinforcement Learning
Post-training
Agentic systems

About the Company

medraai develops Physical AI and robotic systems that operate laboratory instruments alongside biopharma partners to accelerate scientific discovery. The company shipped its initial production system over a year ago, raised $52M in Series A funding, and is establishing a large autonomous laboratory in the United States.

How to Apply

Interested candidates should apply on the official company careers page.

Skills and tech stack

PythonPyTorchJAXLLMsReinforcement LearningDeep LearningMachine Learning

Questions candidates ask

Where is the Research Engineer, Post-training position located?

The job is located in San Francisco.

What programming languages and frameworks are required?

Proficiency in Python is required, along with familiarity with at least one deep learning framework such as PyTorch or JAX.

What does medraai build?

medraai builds Physical AI Scientists, which are robotic systems and scientific AI designed to work with biopharma partners to accelerate scientific breakthroughs.

What is the funding status of the company?

medraai recently raised a $52M Series A funding round.

Ready to apply?

The application is on the official medraai careers page.

Apply on company site