SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Mercor is an AI data company on a mission to organize human intelligence to power the AI economy. The company operates a platform where millions of domain experts train frontier AI models, generating over $4 million per day in expert compensation. Mercor is a profitable Series C company valued at $10 billion, with offices in San Francisco, NYC, and London.
The Applied AI organization builds the systems that convert human expertise into training data for frontier models, including task pipelines, expert workflows, evaluation infrastructure, and supporting services. These systems run synthetic pipelines and modular quality control to generate high-quality tasks at scale, all backed by reliable, fast, and observable backend infrastructure that grows monthly.
As Tech Lead for Applied AI Backend Systems, you will own core services in this stack: designing data models, building APIs and services, and managing data pipelines that move work through the platform. This is a hands-on build role where you take loosely scoped problems, make design decisions, ship to production, and own the outcome. You will mentor other engineers and grow the team.
Key responsibilities include:
- Owning the architecture of the Applied AI backend domain: core services, data models, orchestration systems, and the pipeline execution layer
- Setting technical direction while remaining hands-on enough to build the hardest parts yourself
- Taking undefined problems, deciding what's worth building, scoping the work, writing designs, shipping code, instrumenting it, and maintaining it in production
- Building and tuning high-throughput data and job pipelines with queuing, batching, idempotency, retries, and backpressure; profiling hot spots and managing token/cost attribution
- Owning the design review bar for backend work across the org and mentoring senior engineers
- Provisioning and managing infrastructure as code using Terraform at scale, managing tens of thousands of containers and sandbox environments
- Participating in on-call for owned systems, debugging production incidents, and writing RCCAs
- Driving cross-functional alignment with product, operations, and research partners to translate ambiguous requirements into shipped systems
Requirements:
- 8+ years of professional backend engineering experience building and operating production systems, with a track record of owning architecture across multiple teams and making decisions that aged well
- Experience mentoring senior engineers, not just junior ones
- Strong fundamentals in backend engineering: data structures, algorithms, concurrency, and writing clear, maintainable code
- Hands-on API design experience (REST, gRPC, or GraphQL) with understanding of versioning, contracts, and backward compatibility
- Solid database skills: relational data modeling, indexing, query performance, transactions, isolation, and safe migrations; familiarity with at least one NoSQL or key-value store
- Deep, hands-on expertise in distributed systems: queues and event streams, caching, idempotency, rate limiting, and designing for partial failure
- Experience with data orchestration and workflow management systems like Airflow, Temporal, or Dagster
- Ability to dig into lower-level infrastructure issues: containers, permissions, logs, traces
- Experience running services in production: containers, CI/CD, monitoring, alerting, and debugging under real traffic
- Comfort with ambiguity; ability to take loosely defined problems, ask the right questions, and develop a plan
- Genuine excitement for agentic development and new technology; fluency with modern AI dev tools (Claude Code, Cursor, Copilot); passion for writing good code
- Clear written and verbal communication; high ownership, pragmatism, and bias toward shipping
Nice to have: Experience building or integrating LLM-backed services in production (evaluation, orchestration, or serving); familiarity with AI infrastructure providers like Modal, Fireworks, Baseten, or Temporal.
