Описание: SNI is an IT Outsourcing partner serving prestigious firms and delivering projects worldwide.
Задачи
- Build and maintain Python (FastAPI) services for an AI-driven platform
- Design and implement RAG pipelines covering ingestion, retrieval, generation, and outputs
- Develop document processing workflows for PDF extraction, OCR, and chunking
- Implement hybrid search using vector search and BM25 with tools such as Azure AI Search or Qdrant
- Integrate LLM APIs and manage prompt engineering with structured outputs
- Support frontend integration with Node.js and React and internal tools with Streamlit
- Develop event-driven and asynchronous processing with Celery and messaging
- Ensure code quality, testing, and Docker-based deployments
- Collaborate with business stakeholders to iterate on AI solutions.
Требования
- Strong Python experience, including FastAPI, asynchronous processing, and clean architecture
- Experience with RAG and LLM pipelines and prompt engineering
- Hands-on experience with OpenAI API, embeddings, and structured outputs
- Experience with vector databases such as Qdrant and pgVector and search tools such as Azure AI Search
- Knowledge of document processing, including PDF, OCR, and chunking strategies
- Experience with PostgreSQL and MongoDB
- Familiarity with event-driven architecture, including Celery, RabbitMQ, and Service Bus
- Basic Node.js and React knowledge for integration
- Experience with Docker, CI/CD, and testing
- Strong problem-solving and communication skills
- Nice to have: Experience with LangChain or multi-agent frameworks, knowledge of RAG evaluation and LLMOps, experience with Azure cloud environments.
Условия
- Contract engagement for 12 months with extension
- Start date: ASAP
- Fully remote with occasional quarterly team sessions in Czech Republic