Implement a Vector Similarity Search: from scratch, in code, spelled out.

Let’s build a Vector Search: from scratch, in code, spelled out.

In this session, we will implement the key piece of every RAG system: vector search. We will start with the fundamental building block of every LLMs: embeddings. We will learn what they are, how they can be visualized and what mathematics underly them.

From this udnerstanding, we will move on to implement our own vector search in Java as a custom prodedure for Neo4j. We will see how we calculate the similarity between two vectors, also known as score, and well find the most similar nodes among a list of nodes, also known as the nearest neighbours. Alls this will be implemented in simple and easy to follow code.

Chris Cook is the Co-Founder of Flyweight.io, a young software startup dedicated to developing a new graph-powered Data Collaboration Platform that aims to change the way companies work with their data. With a genuine passion for software engineering, Chris approaches their work with meticulous attention to detail, caring for the platform's technical excellence. Alongside a deep love for all things digital, Chris is known as a lifelong learner, always curious about the latest trends and technological advances.

Chris Cook is the Co-Founder of Flyweight.io, a young software startup working at the interception of Graphs and AI. They use the expressiveness of Graphs to capture meaningful information and find the most relevant pieces for any question. This forms the basis for their AI assistant.

Chris Cook is the Co-Founder of Flyweight.io, a young software startup working at the intersection of Graphs and AI. At Flyweight.io, they leverage the expressive power of graph databases to capture meaningful information and identify the most relevant data for any query. This innovative approach forms the backbone of their AI assistant, enhancing its capability to deliver precise answers.