Description

Design and build REST endpoints in FastAPI with Pydantic v2 request/response models, keeping client contracts stable and backward compatible.
Build and maintain MongoDB data layers (pymongo/motor): schema design, indexes, migrations and backfill scripts, optimistic concurrency, and staying inside document size limits.
Extend the AI chat system: the tool-calling loop on the Anthropic API, MCP tool servers, orchestrator and sub-agents, SSE streaming, and persisted conversations and feedback.
Integrate with Teamcenter (SOA): BOM expansion, session handling, progress reporting, and keeping blocking calls off the event loop with dedicated worker pools and backpressure.
Build and tune the drawing-validation rule engine (60+ rules): geometry and dimension checks, note and material verification, and PDF processing (PyMuPDF, numpy, networkx).
Maintain authentication and authorization: OIDC, Kerberos/SPNEGO, LDAP group lookups, API keys, session cookies and impersonation.
Manage scheduled validation jobs (APScheduler), background workers and notifications.
Write pytest suites (unit, contract and HTTP-level tests; parallel runs with xdist) and keep pre-commit checks passing.
Help with the containerized deployment (Docker, Gunicorn/Uvicorn, nginx, supervisord), the CI/CD pipeline, Sentry monitoring and structured logging with redaction.
Strong Pydantic v2 and typing skills; designs APIs contract-first.
Solid asyncio: can tell when code blocks the event loop, and knows thread pools, timeouts and cancellation.
Production MongoDB experience: data modeling, indexes, aggregation, concurrency control and migrations.
Enterprise authentication: OAuth2/OIDC, and ideally Kerberos/SPNEGO and LDAP.
Testing discipline with pytest: fixtures, dependency overrides, and tests that are safe to run in parallel.
Docker, Linux, reverse proxies (nginx) and running WSGI/ASGI servers (Gunicorn/Uvicorn).
Can write clear design notes and explain trade-offs.

Nice to have
Building LLM applications: the Anthropic or OpenAI SDKs, tool calling, the Model Context Protocol (MCP), streaming, and evaluating prompts.
Siemens Teamcenter SOA or other PLM/ERP integrations. Siemens NX / NXOpen or CAD geometry work.
PDF and image processing (PyMuPDF, Pillow) and numerical Python (numpy, pandas, networkx).
Job scheduling (APScheduler), S3-compatible object storage, Sentry.
Manufacturing or mechanical-engineering domains (GD&T, drawing standards, flex circuits).
Comfortable using AI coding assistants as part of the workflow

Key Skills

Education

Bachelor's degree