Job Description :

Shape the Platform :

- Sit with business, product, and clinical leaders to reframe ambiguous problems into something concrete, scoped, and buildable.

- Validate the riskiest assumption first on every new loopprototype, react, decide what survives.

- Carry architecture and trade-off conversations with stakeholders directly.

- Demo live without a slide deck.

- Lead POCs, innovation sprints, and research experiments to validate emerging AI techniques before they get baked into platform decisions.

- Author ADRs and scope memos for major decisionsLLM abstractions, retrieval design, vendor selection, integration boundaries, infrastructure path.

Architect & Build :

- Own end-to-end architecture for AI-powered platforms : retrieval, reasoning, evaluation, integration, and the seams between them.

- Design vendor-agnostic LLM abstractions so frontier models (Claude, GPT, Gemini, open-weight) can be swapped behind a clean interface as enterprise constraints evolve.

- Architect and ship production-grade agentic systems using LangGraph, AutoGen, Claude Agent SDK, OpenAI Assistants, or your own orchestration layer.

- Build modular backends in Python or TypeScript aligned with clean architecture, OOP, SOLID, and domain-driven design.

- Apply RAG techniques where they actually help : vector databases (Pinecone, Chroma, Weaviate, pgvector), hybrid retrieval with ElasticSearch or Solr, BM25 + similarity, re-ranking.

- Design prompt and context engineering frameworks that optimize accuracy, repeatability, cost, and latency.

- Use AI-assisted development tools (Claude Code, Cursor, GitHub Copilot, Codex) through structured workflows, sub-agents, skills, and templateswith discipline and review.

Productionize & Operate :

- Spin up the infra, write the evals, wire the MCP servers, deploy the agents, and harden the bits that survive contact with real users.

- Deploy on AWS (or Cloudflare for edge use cases) using containerization (Docker, Kubernetes, ECS) or serverless (Lambda)chosen for fit, not preference.

- Treat evals as a first-class discipline : hands-on harnesses, golden datasets, regression rubricsnot theoretical frameworks.

- Apply engineering practices that hold up in production : TDD, secrets management and rotation, SAST/DAST, audit trails, RBAC, structured logging, metrics, tracing, automated CI/CD (GitHub Actions, Jenkins).

- Engage enterprise architecture review paths early when they apply.

- Move with them, not around them.

- Mentor others on system design, agentic patterns, and AI engineering best practices.

What You Bring :

- 8+ years engineering experience, with architecture-level ownership on at least one production system.

- Direct hands-on experience designing and shipping LLM-powered products end-to-end : RAG pipelines, prompt and context engineering, eval harnesses, vendor-agnostic LLM abstractions.

- Hands-on experience with agents, not just prompted models.

- You have wired tools to a model and let it run multi-step using LangGraph, AutoGen, Claude Agent SDK, OpenAI Assistants, or your own orchestration.

- Strong Python or TypeScript, with OOP, SOLID, 12-factor application development, and microservice architecture.

- You have built Next.js applications, FastAPI services, and similar.

- End-to-end implementation experience with vector databases, retrieval pipelines, and eval harnesses.

- Cloud-native AWS deployment experiencewith Docker, Kubernetes, and GitHub Actions.

- Cloudflare experience a plus.

- Active, structured use of AI-assisted development tools (Claude Code, Cursor, GitHub Copilot) with demonstrable workflows, sub-agents, skills, and templates.

- A deep working understanding of how LLMs behaveand where they breakand how to optimize for accuracy, latency, and cost.

- Track record working with evolving requirements and co-creation models, not finished specs.

- Strong written communicationyou author ADRs, scope memos, and decision documents that hold up under review.

- Comfortable making trade-off calls in front of business leadership without locking them in.

- A real, recent trail of built things : GitHub, a portfolio, side projects, indie tools, or OSS contributions.

- A no-compromise attitude on clean code, TDD, security, observability, scalability, performance, and cost.

- A founder's mindset and genuine appetite for ambiguous, high-impact technical challenges.

- Bachelor's or Master's in Computer Science, Machine Learning, or a related technical discipline.

- Public writing, talks, or threads about building with AI.

- MLOps and model serving experience (BentoML, MLflow, Vertex AI, SageMaker).

- Streaming and batch ingestion pipelines (Spark, Airflow, Beam, Glue).

- Experience with enterprise AI governance frameworks (EU AI Act readiness, internal AI policy / risk frameworks).

- Healthcare or life sciences domain exposure.

- Pharma, healthcare, or other regulated-industry experience.

- Relevant cloud architecture certifications