AI / LLM Engineer
DEVAPO SPÓŁKA Z OGRANICZONĄ ODPOWIEDZIALNOŚCIĄ
⚲ Warszawa
Wymagania
- Python
- Docker
- AWS
- Microsoft Azure
- Google Cloud Platform
- LangChain (nice to have)
- LangGraph (nice to have)
- LlamaIndex (nice to have)
- Qdrant (nice to have)
- Weaviate (nice to have)
- Pinecone (nice to have)
- pgvector (nice to have)
- RAGAS (nice to have)
- LangSmith (nice to have)
- Arize (nice to have)
- Databricks (nice to have)
- Azure AI Foundry (nice to have)
- AWS Bedrock (nice to have)
Opis stanowiska
Wymagania: - You've built something with LLMs that actually runs in production — doesn't matter if it was at a big company or a side project - Solid Python skills — not scripts, but clean code you're not ashamed of - Working knowledge of RAG: you understand why naive chunking fails and what to do about it - Experience with at least one agent framework (LangChain, LlamaIndex, LangGraph) - Familiarity with vector databases (Qdrant, Weaviate, Pinecone, pgvector — any) - You know how to ship: REST APIs, Docker, cloud basics (AWS / Azure / GCP) - English B2+ — client-facing role, calls and written communication included Mile widziane: - Experience evaluating LLM outputs (RAGAS, LangSmith, Arize or similar) - MLflow or another experiment tracking tool - Databricks, Azure AI Foundry or AWS Bedrock - Fine-tuning experience (LoRA, PEFT, anything hands-on) - Kafka or streaming pipelines for real-time AI use cases O firmie: - We provide software engineering and data specialists to international clients who want people, not vendors. Our engineers work embedded in client teams on real problems — not slide decks. We invest in your growth: certifications, learning time, architecture access, and honest feedback from people who know what they're talking about. - If you like figuring things out and shipping them — you'll fit in. Zakres obowiązków: - Building LLM-powered applications and RAG systems for enterprise clients - Designing and implementing AI agents (LangChain, LangGraph, CrewAI or similar) - Integrating LLMs (OpenAI, Anthropic, Gemini, open-source) into existing systems - Building data ingestion pipelines: chunking, embedding, vector indexing - Writing production-grade Python code — APIs, tests, containers, the full stack - Working directly with clients: understanding their requirements, presenting solutions - Doing code reviews, writing docs, contributing to team engineering standards Oferujemy: - Certifications and training funded - Private medical care (Medicover) - Multisport card - English language classes - Flexible working hours - Team meetups and integration events - Referral bonus