CareerPlan

Research, Coding Agents

San FranciscoFull-time$400k–$500k2026-08-27 → 2026-11-19
On-siteVisa sponsorship offered

Core

Research role focused on building AI models capable of agentic coding, including writing, debugging, and reasoning about code across long-horizon tasks.

Role type

Senior Research Engineer (Agentic Coding)

Builds

Frontier AI models with world-class agentic coding capabilities

Domain

Artificial Intelligence / Machine Learning / Software Engineering

Deliverable

production ML models

Required skills

Python, Deep Learning frameworks (PyTorch, TensorFlow, JAX), RL training, synthetic data generation, reward design, distributed training debugging

Preferred skills

End-to-end model usability gap identification, large-scale agentic RL infrastructure, frontier model coding capability improvement

Technologies

PyTorch, TensorFlow, JAX, RL environments, sandboxes

Responsibilities

Design and run RL training jobs for agentic coding, build sandboxed coding environments, generate synthetic coding data, design evaluations for coding usefulness, debug large RL runs

Seniority

Senior, hands-on IC