AI Resident
Core
You own one hard problem end-to-end in production Agentic AI, building systems that run real business processes at scale.
Role type
AI Resident (Research & Engineering)
Builds
Self-improving agent harnesses, inference-cost optimization, agent memory, and evaluation infrastructure.
Domain
Enterprise productivity, Agentic AI, Production ML
Deliverable
production ML models
Required skills
Python, PyTorch, Post-training (SFT/DPO/RL), Reward modeling, Agent and tool-use systems, Statistical literacy, Honest measurement
Preferred skills
Hands-on post-training with open models, Interactive agent environments, Large-scale trace analysis, Serving and efficiency (vLLM/SGLang), Multi-node GPU training
Technologies
PyTorch, vLLM, SGLang, TRL, veRL, OpenRLHF
Responsibilities
Write proposals and build systems end-to-end, Design evaluation metrics and run experiments, Ship code into production codebase with senior mentorship, Analyze production traces for training data, Optimize inference costs and agent efficiency
Seniority
Junior to Mid-level, hands-on IC
