Stavros Kontopoulos is a software engineer with many years of experience in crafting diverse systems and products. He loves contributing to Open Source and he is passionate about distributed system design, data engineering, eventing architectures and ML & AI technologies.
Building a chatbot powered by LLMs and tailored to your own data is one of the most practical and in-demand AI applications today. In this talk, we’ll demystify the process of building one from scratch using Python, breaking it down step by step.
We’ll cover:
Project setup – Laying the foundation for a scalable chatbot
Best practices – Structuring an AI project effectively
Data quality – Techniques to improve and preprocess your data
Enhancing responses – Leveraging Retrieval Augmented Generation (RAG), Cache Augmented Generation (CAG), and agentic RAG for smarter interactions
Evaluation – Key metrics for assessing chatbot performance
Fine-tuning – When (and when not) to customize your model
Guardrails – Implementing safeguards for input and output control
Deployment – Running your chatbot on Kubernetes for seamless orchestration
By the end of this session, you'll have a solid understanding of the essential components needed to build, refine, and deploy your own AI-powered chatbot. Whether for fun or profit, this is your roadmap to making it happen.
Searching for speaker images...