Founding AI Research Engineer: Agent Harness & Model Systems
Core
Building an agentic runtime (harness) for AI agents that manage the entire hiring and talent lifecycle, including career strategy, matching, and compliance.
Role type
Founding AI Research Engineer (Agent Systems & Model Evaluation)
Builds
An orchestration runtime for multi-agent systems, proprietary domain models for the hiring domain, and a compliance-auditable trace system.
Domain
AI Agents, Hiring Technology, Compliance (EU AI Act, FCRA)
Deliverable
production ML models
Required skills
Multi-agent orchestration, tool-use and function-calling, context engineering, model routing, in-harness evaluation, compliance instrumentation, corpus curation
Preferred skills
Open-weight model evaluation (Qwen, Mistral, DeepSeek), agentic fine-tuning, schema validation, failure taxonomy design
Technologies
Qwen, Mistral, DeepSeek, Python, LLM frameworks
Responsibilities
Design and build the orchestration runtime (planner/executor, state management, handoffs); develop tool-call contracts and guardrails; engineer context assembly and retrieval policies; evaluate and select base models for the harness; build the data engine for corpus curation; implement auditable traces for regulatory compliance.