Founding Senior Applied AI Engineer

Sourced from jobs.ashbyhq.com

Goaly is a stealth AI startup founded by ex-Meta Superintelligence Labs engineers, building infrastructure to help every company become its own AI lab. As a founding Senior Applied AI Engineer, you'll build the agent harness layer between frontier models and real users — the agent loop, tools, context engineering, sandboxes, and evals. You'll ship to production for real customers and feed learnings back into better environments, evals, and models.

• Own the agent harness: loop, tool interface, system prompts, permissions, stop conditions, retries, budget caps, and output checks
• Build sandboxed execution environments (containers/microVMs) for parallel rollouts used in serving, RL, and evals
• Design context engineering: window contents, compaction, file-backed memory, retrieval, and sub-agent hand-offs
• Make agent runs observable and replayable by tracing model calls, tool calls, and state changes
• Build the eval stack: offline suites, production-trace tests, LLM-as-judge, and reward-hacking checks
• Ship with customers from prototype to stable production and turn patterns into reusable building blocks
• Close the loop with research by converting production traces and failures into tasks, environments, and reward signals
• Own model selection and cost, latency, and quality trade-offs

• 5+ years building production software with experience running systems at scale
• Built and shipped an agentic system with tool use that real users depend on
• First-hand experience with agent failure modes: context overflow, looping, tool misuse, prompt injection, runaway cost
• Strong Python plus at least one of TypeScript, Go, or Rust
• Comfortable with Docker, Linux, and cloud infrastructure
• Habit of building custom evals rather than trusting public benchmarks
• Comfort operating in ambiguity without a playbook
• Ability to explain trade-offs clearly to engineers, researchers, and customers

• Experience building agent platforms, workflow systems, model-serving products, or ML infrastructure
• Experience with multi-tenant control planes, IAM, metering/billing, enterprise security, or high-throughput APIs
• Familiarity with LLM inference, fine-tuning, evaluation, tool use, sandboxes, or production agent lifecycle
• Experience owning open-source SDKs, developer tools, API docs, or community integrations
• Full-stack ability and interaction design sense