Freelance AI Consultant | LLM & Knowledge Graph Specialist | RAG Systems Developer

Are you looking to unlock the true potential of your complex data using cutting-edge Generative AI? As a Freelance AI Consultant with a Ph.D. in Theoretical Physics, I specialize in designing, developing, and implementing bespoke Retrieval-Augmented Generation (RAG) systems, particularly those powered by the synergy between Large Language Models (LLMs) and robust Knowledge Graphs (KGs).
My core expertise lies in bridging the gap between structured domain knowledge and the generative power of LLMs. I architect solutions that go beyond simple vector search, leveraging graph databases like Neo4j to build sophisticated Knowledge Graphs. This allows for more nuanced, context-aware, and accurate information retrieval, forming the backbone of powerful GraphRAG systems. I possess deep experience in:

• Knowledge Graph Design & Implementation: Crafting effective ontologies and graph schemas, modeling complex relationships within data, and proficiently using Neo4j (Cypher querying, database management, Graph Data Science library).
• LLM Integration & RAG Pipelines: Integrating various LLMs (via APIs like OpenAI or using open-source models), developing robust RAG architectures, optimizing retrieval mechanisms, and proficient prompt engineering for question-answering and conversational AI.
• GraphRAG Implementation: Specific experience in designing and building RAG systems where the knowledge graph is central to retrieving interconnected and contextual information, significantly enhancing the quality and relevance of LLM responses.
• End-to-End PoC & Prototype Development: Rapidly developing proof-of-concepts and functional prototypes to validate GenAI approaches for specific business needs, particularly within complex domains like R&D.

Proven Experience:
During my time as a Data AI Engineer at Capgemini Engineering, I played a key role in developing impactful GenAI solutions:

• Sanofi Bioproduction Knowledge Graph & RAG System: Engineered a sophisticated RAG system using Neo4j and LLMs. This involved modelling extensive experimental data and enabling scientists to query complex information (like batch genealogy) via a natural language chatbot, significantly improving research workflow efficiency and accelerating data-driven decision-making.
• RATP Rail Operations Chatbot (PoC): Designed and prototyped a conversational AI applying RAG principles over a Neo4j graph, demonstrating the potential to simplify access to complex operational data (schedules, incidents) via natural language.
• Agentic AI Systems (CrewAI): Explored advanced AI architectures by developing multi-agent systems integrated with knowledge graphs (PostgreSQL-based KG in this instance) to tackle complex engineering problems, showcasing adaptability with cutting-edge tools.

Technical Skillset:
Beyond my core focus, I am proficient in Python (Pandas, NumPy, Scikit-learn, etc.) for data processing, analysis, and ML model development (Regression, Classification, etc.). I have familiarity with TensorFlow for neural networks and experience with tools like CrewAI for multi-agent systems. My background includes extensive research, resulting in peer-reviewed publications and demonstrating strong analytical and problem-solving capabilities honed during my Ph.D.
Seeking Opportunities:
I am actively seeking freelance projects focused on developing innovative RAG systems (especially GraphRAG), LLM-powered applications, and knowledge graph solutions. If you need expertise in transforming your data into actionable insights and intelligent applications, let's connect and discuss how I can contribute to your success.