Senior AI Data Engineer (LLM / RAG)

SouthRivers Data is an EU-based data and AI engineering company headquartered in Bratislava, Slovakia. We design and deliver data platforms and production AI systems for companies in the US, UK and Europe, and run several client projects in parallel.

Our client, a US-based B2B company, is building an AI-powered analytics platform on top of its sales and operational data. The platform ingests data from the CRM and other business systems, models it in PostgreSQL and calculates business metrics in a dedicated analytics layer. Teams then explore the results through a natural-language interface. The LLM works only on validated data and calculated results, so every number it reports can be traced back to the source.
You join as the senior engineer responsible for the data and AI layers of the platform.

Your responsibilities: 

• Build ingestion pipelines and API integrations with the CRM and other business systems
• Design the PostgreSQL data model and the metrics layer
• Build the LLM layer: natural-language querying through tool calling, and classification of call transcripts and other unstructured data
• Implement RAG over company documents with hybrid search and re-ranking
• Deploy and run the platform in the cloud with tests, monitoring and cost tracking
• Present architecture decisions and demos to the client

What we expect:

• 5+ years in data or software engineering, including 2+ years building LLM features in production
• Strong Python (FastAPI, pandas/polars, pytest) and advanced SQL
• PostgreSQL data modelling and migrations on live data; experience with BigQuery or Snowflake
• Pipeline orchestration with Airflow, Prefect or dbt, including incremental loads, backfills and data-quality checks
• API integrations: authentication, pagination, rate limits, retries, changing source schemas
• LLM APIs (OpenAI, Anthropic, Gemini): structured outputs, tool calling, cost tracking, evaluation of answer quality
• At least one RAG system shipped to real users: chunking, embeddings, pgvector or a vector database, hybrid search, re-ranking
• Deployment on GCP or AWS with Docker, CI/CD and proper secrets management
• Daily work with Claude Code or a similar AI coding agent, and the judgement to review its output critically
• English at Upper-Intermediate or higher; you will lead client calls and demos

Nice to have:

• TypeScript/React or Next.js for internal tools
• LangGraph, Claude Agent SDK, Langfuse or LangSmith
• Kafka or other streaming systems
• Knowledge graphs, entity resolution, embedding-based clustering
• Experience with marketing, sales or CRM data
• Direct work with US clients

What we offer:

• Work on production AI systems for clients in the US, UK and EU
• Continuous engagement across projects, with no gaps between assignments
• Fully remote with a flexible schedule
• Possible part-time collaboration
• Architecture reviews and technical guidance from a Data Architect
• Real ownership: you make architecture decisions and present them to the client
• Long-term collaboration

Beyond this project:
We run multiple data and AI projects at the same time: RAG systems, LLM processing pipelines, data platform modernization, and technical audits of AI platforms. Depending on project priorities, you may move to another project. Any move will use a similar stack, match your seniority, and come with advance notice. When one engagement ends, the next one is already in place.