Senior Data Scientist with Graph

It is an opportunity for a Senior Data Scientist working on advanced AI initiatives, turning research ideas into production-ready AI products. The role combines core ML/Data Science with Graph AI, Knowledge Graphs, NLP/LLMs, RAG and agentic workflows.

ROLE / PROJECT

Build end-to-end AI solutions across multiple domains, combining structured and unstructured data, graphs, LLMs and enterprise services. You will work on graph and embedding pipelines, knowledge graphs, retrieval/RAG, agentic workflows and production ML systems, taking solutions from experimentation and prototyping through deployment and monitoring.

THE SETUP

Type of cooperation: Contract Location: 100% remote Working hours: Required 10 AM - 6 PM CET, ideally 11 AM - 7 PM CET Start: ASAP Language requirements: English

TECH STACK

Python, ML/DL, LLMs, RAG, NLP, Knowledge Graphs, Neo4j, GNN (GCN/GAT), embeddings, vector/hybrid search, text-to-SQL, APIs, microservices, Azure/AWS/GCP, CI/CD

MUST HAVES

• Strong ML, Deep Learning and Data Science fundamentals
• 3-6+ years of strong Python engineering experience
• Experience with Graph AI / Knowledge Graphs and/or graph data science
• Strong NLP/LLM knowledge, including RAG and semantic engineering
• Retrieval/search systems: lexical, vector or hybrid search, embeddings and reranking
• Hands-on experience with RAG, agentic workflows or similar AI patterns
• Model evaluation, experimentation, testing and observability
• APIs, microservices and data-centric integrations
• Cloud experience (Azure, AWS or GCP) and CI/CD
• Strong debugging, profiling and performance optimization skills
• Git, code reviews and agile delivery
• Availability to work 10 AM - 6 PM CET, ideally 11 AM - 7 PM CET
• HackerRank Challenge as part of the selection process
• Freelance / must have own company or trade licence.
• English B2+

NICE TO HAVES

• Neo4j and Knowledge Graph modeling
• GNN experience, particularly GCN/GAT
• Fine-tuning, personalization and text-to-SQL
• GPU / accelerated training and inference
• Distributed computing such as Ray, Spark or Dask
• Experience optimizing AI workloads for performance and cost