Python Engineer (Remote)

As Senior Python Engineer you are the second-most-senior engineer on the AI track. You are not expected to lead LangGraph architecture, but you must be able to read it, debug it under production load, and ship safe changes around it. Your primary ownership is the FastAPI surface, the Celery topology, and the data layer.

This is a hands-on individual-contributor role with growth into a tech-lead role if you want it.

What you'll do

First 90 days

• Onboard onto the FastAPI chat-api and Celery flows alongside the AI Tech Lead
• Audit the FastAPI surface: routes, dependencies, middleware, streaming, error handling
• Audit the Celery topology: queues, workers, retry policy, scheduled jobs, dead-letter handling
• Profile and document the multi-datastore patterns (Mongo + Redis + Postgres) and any race-condition risks
• Help build evaluation and regression test infrastructure

Ongoing

• Own the FastAPI service health (response time, error rates, streaming reliability)
• Own the Celery operational layer (queue lag, worker health, dead-letters)
• Implement new endpoints and async flows alongside agent changes from the AI Tech Lead
• Participate in on-call rotation for the AI core
• Pair with mid-level Python engineers as the squad grows

Must-have skills:

• 5+ years Python in production
• Production FastAPI experience — async routes, dependency injection, middleware, streaming responses, OpenAPI generation
• Celery + Redis broker — building and operating task queues at scale, understanding retry semantics, queue routing, scheduled jobs
• Async Python — asyncio depth, structured concurrency, knowing when async helps and when it hurts
• MongoDB in production (PyMongo or Motor)
• Postgres in production — schema design, indexing, transactions, basic query tuning
• Redis as both cache and broker
• Pydantic v2 fluency, type-hinted code, mypy or pyright comfort
• pytest including pytest-asyncio
• Production debugging — logs, traces, profiling, identifying memory leaks
• Working English

Nice-to-haves:

• LLM API exposure (OpenAI, Anthropic) at consumer level — you don't need to lead AI work, but you should be able to follow it
• LangChain awareness (not LangGraph depth)
• Data engineering or ETL background
• Datadog APM with Python
• Knowledge of fitness / health / wearables domain
• Experience taking over a legacy codebase
• Vector store exposure (pgvector, Pinecone)