Senior Software Engineer, AI

What You'll Do

• Design and ship LLM-powered product features end to end, including prompt and context design, structured output, tool/agent orchestration, and the UI/API work needed to land them in the product.
• Pair with product-vertical engineers to unblock AI feature work and help teams apply August Intelligence patterns in their own product areas.
• Turn repeated AI implementation work into reusable tools, examples, documentation, and lightweight practices that other teams can adopt.
• Build and extend our evaluation discipline so we can measure quality, catch regressions, and reason about nondeterministic behavior before customers feel it.
• Improve observability for AI workflows, including traces, prompt visibility, errors, latency, cost, and production feedback loops.
• Harden AI features for production: latency, cost, caching, pre-computation, timeouts, cancellation, memory behavior, and operational reliability.
• Handle PHI and sensitive resident data with care, including redaction, permissioning, tenancy boundaries, and safe access patterns.
• Partner with product managers and vertical engineering teams to translate clinical and operational workflows into AI capabilities.
• Help August learn where AI can make the product meaningfully more useful by prototyping, measuring, and refining ideas against real workflows.
• Help non-engineers experiment safely with real data while creating a clear path from prototype to production.
• Help drive responsible AI tool use inside engineering by modeling effective workflows, reviewing AI-assisted output, and raising shared standards without turning experimentation into ceremony.
• Review code, including AI-assisted code, with rigor and help set August Intelligence's engineering standards.

About You

• 5+ years of experience building and shipping production software.
• Owned customer-facing features end to end and care about what happens after launch.
• Strong in production TypeScript, including typed Node services, testing, APIs, and clean data modeling.
• Hands-on applied-LLM experience: prompt design, structured output, tool or agent design, and practical reasoning about cost, latency, hallucination, and reliability tradeoffs.
• Bias toward the smallest useful thing: ship a walking skeleton, learn from it, and iterate rather than guessing future states and going big too early.
• Can explain exactly how your code works, including code an LLM helped you write.
• Enjoy pairing with other engineers, teaching through real work, and helping teams adopt new patterns without taking ownership away from them.
• Comfortable working with sensitive data and treat permissioning, privacy, and safety as table stakes.
• Communicate clearly in writing and conversation, especially when working with PMs, clinicians, operators, and engineers from other product areas.
• High agency and low ego. Like shaping ambiguous problems, not just implementing tickets.

Nice to Have

• Experience with Mastra, the Vercel AI SDK, LangChain, LlamaIndex, or similar applied-LLM frameworks.
• Experience with eval or observability tooling for LLM systems, such as Phoenix/Arize, Langfuse, Helicone, or related tools.
• Experience helping an engineering team adopt AI tools, agentic coding workflows, or new development practices responsibly.
• Experience with functional TypeScript or Effect.
• Ability to read Scala or interest in learning enough Scala to navigate our backend.
• Healthcare, senior living, SOC 2, HIPAA, or other regulated-industry experience.
• Experience as an early engineer on a small, high-ownership team.