You will be redirected to djinni.co to apply.
About the project
US-founded media company operating email newsletters at scale - millions of subscribers across finance, health, and business verticals. Small engineering team, flat structure, direct ownership, fast iteration. No layers of approval, no large org to navigate.
You'll own the AI layer of the automation platform alongside a counterpart engineer who handles integration and infrastructure.
What you'll work on
- AI-driven content generation and management workflows at production scale
- Agent systems coordinating multi-step operational tasks: content, QA, validation, reporting
- RAG pipelines over internal knowledge bases and operational data
- Prompt architectures designed for reliability, cost control, and measurable output quality
- Evaluation and regression testing for LLM systems - "it looked fine when I tested it" isn't good enough
Real production systems, not demos. The work removes hours of manual operational work daily and needs to be reliable enough that people actually trust it.
Requirements
- 6+ years of development experience
- 4+ years of production Python
- 2+ years shipping production systems with LLMs and agents
- Solid understanding of what happens underneath the frameworks: tokenization, context windows, function calling, when RAG helps and when it doesn't
- Hands-on with agent orchestration: LangChain, LangGraph, MCP, or equivalent
- Clean, maintainable Python - frameworks support your thinking, not replace it
Tech stack
Python · FastAPI · PostgreSQL · LLM APIs (OpenAI, Anthropic) · LangChain / LangGraph / MCP · vector databases · Playwright · Docker · proxy infrastructure
Conditions
- B2B contract, full-time
- Warsaw office (onsite preferred, remote considered)
- Direct collaboration with technical PM and small senior team
- Fast-moving environment: ambiguity, shifting priorities, growing scope