SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Mercor is an AI data company building infrastructure between human expertise and frontier AI models. The company operates a platform where millions of domain experts train frontier AI models, and is expanding into enterprise with agentic systems that capture and replicate expert workflows.
As a Member of Technical Staff on the Talent Experience team, you'll own the backend systems and agent infrastructure powering Mercor's talent network. This is production agent engineering at scale—you'll design and ship services that orchestrate agent workflows, manage tool interfaces, implement evaluation and guardrail layers, and maintain data models that keep agent decisions auditable and reversible.
Key responsibilities include:
- Design, build, and own backend services powering agentic workflows: orchestration, tool calling, retries, fallbacks, and human-in-the-loop review systems
- Ship agents into production and iterate on them against real usage, optimizing prompts, tools, guardrails, and cost/latency tradeoffs
- Build evaluation infrastructure to measure agent effectiveness: offline evaluations, online metrics, and regression detection
- Design core services, data models, and APIs that scale with the talent network while remaining legible as agent behavior evolves
- Make agent decisions observable, auditable, and reversible to catch and understand failures quickly
- Partner with product, operations, and AI researchers to determine what should be agentic, what should remain deterministic, and where humans belong in the loop
- Establish backend engineering standards and patterns, raising the technical bar across the team
You'll need 5+ years of professional backend engineering experience with strong fundamentals in distributed systems, service architecture, data modeling, and API design at scale. Hands-on experience building with LLMs in production (agents, tool use, retrieval, evaluation) is required, or a demonstrated ability to learn new systems quickly and ship them. You should be comfortable with ambiguity—agent systems fail differently than traditional services—and have a track record of owning systems end-to-end from design through production operation. Fluency with modern AI development tools (Cursor, Claude Code, GitHub Copilot) is expected. You'll communicate complex tradeoffs clearly to both engineers and leadership, drive alignment through technical judgment, and demonstrate high ownership and pragmatism with a bias toward shipping.
Mercor is a profitable Series C company valued at $10 billion, operating in-person five days a week from San Francisco, NYC, or London offices. This role is SF-only.
