AI Seminar Series - Sam Elamin

In this talk, we will delve into the journey of building a generative AI platform, sharing the insights and lessons learned along the way. We will explore the high-level architecture that underpins the platform, highlighting the key components and their interactions. The discussion will cover the intricacies of prompt engineering, the challenges faced, and the strategies employed to overcome them. We will also touch upon Retrieval-Augmented Generation (RAG), SQL generation, and the role of embeddings and vector databases in enhancing the platform's capabilities.

The session will provide a candid look at what worked and what didn't, offering valuable takeaways for anyone looking to embark on a similar journey. We will discuss the difficulties encountered, from technical hurdles to operational challenges, and how we navigated them. Additionally, we will showcase some of the practical applications built on the platform, such as a bot that can answer questions about subscription and viewing metrics. For instance, the bot can respond to queries like "How did House of the Dragon season 1 perform since launch?" or "What are the top titles this week?" By the end of the talk, attendees will have a comprehensive understanding of the complexities involved in building a generative AI platform and the practical solutions that can be applied to their own projects.