Solely using out-of-the-box large language models (LLMs) for information retrieval leads to inaccuracies and hallucinations as they do not encode domain specific proprietary knowledge about an organization's activities. We will demonstrate how a Memgraph + DeepSeek Retrieval Augmented Generation (RAG) solution provides more “grounding context” to an LLM and obtains more relevant, specific responses.

Robert leads Data Science and Analytics projects for Commercial and Federal financial institutions. He has extensive experience in financial engineering and risk management, with 25 years of experience working for investment banks and hedge funds. Robert is an expert in graph technology which he has used to deliver solutions in data harmonization, entity resolution, and customer experience. Robert holds a Master’s degree in Mathematics of Finance from Columbia University, and a Master’s degree in Information Management Systems from the Harvard Extension School.

Sinan uses his expertise in Graph Databases, Machine Learning, and Large Language Models (LLMs) to offer technical consulting services for FI Consulting’s Federal and Commercial clients. Sinan’s skill in programming languages, such as Python and R, and AWS and Azure products, helps create seamless applications that blend Generative AI with statistical analysis in solving challenging business problems.

With extensive experience across federal agencies and financial institutions, James is a software and data engineer passionate about leveraging emerging technologies to drive transformative change. He has successfully led teams in developing full-stack applications, optimizing data pipelines, and delivering scalable software solutions for large-scale, billion-dollar programs. James has utilized cutting-edge GPU-based machine learning models and Graph Databases to address some of FI Consulting's most complex data challenges.

Memgraph + DeepSeek for Retrieval Augmented Generation (RAG)