AI Engineer, Agents and Evaluation

Pronoia•Torrance, CA

1d•Onsite

About The Position

Pronoia Energy is a deep tech startup developing room-temperature macroscopic quantum energy storage technology. The company is building an internal, grounded agentic AI system to accelerate its research. The AI Engineer will own the orchestration and verification layers: planning research questions into steps, routing steps to appropriate tools, and preventing unsupported claims. The system must route factual questions to retrieval and computational questions to tools, without answering from the model’s own memory. The engineer will also deploy a self-hosted open-weights model on local hardware and build the agent loop around it. This is a founding generalist role with responsibilities spanning orchestration, model serving, and evaluation, with the possibility of handing specialties to dedicated hires later. The role is onsite in Torrance, California.

Requirements

Track record of shipping a real agent or tool-use system in production, rather than a demo.
Hands-on experience self-hosting open-weights models on the candidate’s own GPUs, rather than only using hosted APIs.
Treat hallucination and tool misuse as engineering problems that require mechanical fixes.
Generalist capabilities across orchestration, model serving, and evaluation.
Self-starter who thrives amid ambiguity and rapid iteration.

Nice To Haves

Scientific or physics background.
Contributions to open-source serving or agent tooling, such as vLLM, TGI, or agent frameworks.
Experience building retrieval-augmented systems and evaluation suites.
Prior work in deep tech R&D alongside interdisciplinary teams.

Responsibilities

Build the agent orchestrator and tool-routing logic.
Stand up and serve a self-hosted open-weights model on local hardware.
Build the first verification gate and evaluation harness so that grounding is mechanically enforced.
Ensure claims trace to a retrieved source or a deterministic computation.
Work across orchestration, model serving, and evaluation as a founding generalist, with the expectation of handing specialties to dedicated hires later.