What you'll do:

Architect & build multi-agent systems using frameworks such as LangChain, LangGraph, AutoGen, Google ADK, Palantir Foundry, or custom orchestration layers.
Fine-tune and prompt-engineer LLMs (OpenAI, Anthropic, open-source) for retrieval-augmented generation (RAG), reasoning, and tool use.
Integrate agents with enterprise data sources (APIs, SQL/NoSQL DBs, vector stores like Pinecone, Elasticsearch) and downstream applications (Snowflake, ServiceNow, custom APIs).
Own the MLOps lifecycle: containerize (Docker), automate CI/CD, monitor drift & hallucinations, set up guardrails, observability, and rollback strategies.
Collaborate cross-functionally with product, UX, and customer teams to translate requirements into robust agent capabilities and user-facing features.
Benchmark & iterate on latency, cost, and accuracy; design experiments, run A/B tests, and present findings to stakeholders.
Stay current with the rapidly evolving GenAI landscape and champion best practices in ethical AI, data privacy, and security.

Must-Have Technical Skills

35 years software engineering or ML experience in production environments.
Strong Python skills (async I/O, typing, testing) plus familiarity with TypeScript/Node or Go a bonus.
Hands-on with at least one LLM/agent frameworks and platforms (LangChain, LangGraph, Google ADK, LlamaIndex, Emma, etc.).
Solid grasp of vector databases (Pinecone, Weaviate, FAISS) and embedding models.
Experience building and securing REST/GraphQL APIs and microservices.
Cloud skills on AWS, Azure, or GCP (serverless, IAM, networking, cost optimization).
Proficient with Git, Docker, CI/CD (GitHub Actions, GitLab CI, or similar).
Knowledge of ML Ops tooling (Kubeflow, MLflow, SageMaker, Vertex AI) or equivalent custom pipelines.
Core Soft Skills
Product mindset: translate ambiguous requirements into clear deliverables and user value.
Communication: explain complex AI concepts to both engineers and executives; write crisp documentation.
Collaboration & ownership: thrive in cross-disciplinary teams, proactively unblock yourself and others.
Bias for action: experiment quickly, measure, iteratewithout sacrificing quality or security.
Growth attitude: stay curious, seek feedback, mentor juniors, and adapt to the fast-moving GenAI space.

Nice-to-Haves

Experience with RAG pipelines over enterprise knowledge bases (SharePoint, Confluence, Snowflake).
Hands-on with MCP servers/clients, MCP Toolbox for Databases, or similar gateway patterns.
Familiarity with LLM evaluation frameworks (LangSmith, TruLens, Ragas).
Familiarity with Palantir/Foundry.
Knowledge of privacy-enhancing techniques (data anonymization, differential privacy).
Prior work on conversational UX, prompt marketplaces, or agent simulators.
Contributions to open-source AI projects or published research.

AI Software Engineer (LLM & MLOps)