AI Applications Engineer

Role overview A remote, full-time lead engineering role focused on building and operating AI companions deployed inside other companies. The position owns a single client engagement end to end, covering agents, tools, retrieval, the application layer, backend services, deployment, and ongoing operating cost. It is suited to a full-stack engineer comfortable leading technical decisions and shipping weekly.

Responsibilities - Own a client AI companion end to end, including agents, tools, retrieval pipelines, app surfaces, backend services, deployment, and runtime cost. - Build with Claude and the Anthropic SDK, including streaming tool-use loops, structured outputs, prompt caching, and evaluation harnesses that catch incorrect outputs before users see them. - Develop stateful agents capable of remembering prior context, scheduling follow-up work, and acting autonomously within defined guardrails. - Ground every response in cited sources, ensuring the model never improvises on legal, financial, or procedural questions. - Migrate production data off legacy systems without losing any records. - Ship weekly alongside client product owners and document what was released.

Requirements - Three or more years of experience shipping full-stack software used by real users, including on-call responsibility for systems you built. - Strong backend fundamentals: APIs, relational data, background jobs, containers, and deployments, with the ability to stand a service up and keep it running. - Deep TypeScript proficiency and solid grounding in HTTP, SQL, concurrency, and data structures. - Working knowledge of at least one major cloud platform (Google Cloud, AWS, or Azure) and willingness to learn the others. - Production experience with language models, including prompting, context, tool use, retrieval, evaluation, latency, and cost. - Clear written and spoken English, with the ability to explain technical decisions to non-engineer founders. - Demonstrated fluency with Claude Code and VS Code, with concrete examples of work done in both. - A bachelor's degree in computer science or a closely related field, though exceptional candidates with a strong body of work are also considered.

Nice to have - Experience with edge runtimes or stateful serverless systems. - Background migrating legacy applications without downtime. - Multilingual product work. - Evaluation and safety testing for language-model applications.