How to build a secure, authorized chatbot using Oso and Timescale

As large language models (LLMs) become increasingly popular for powering natural language interfaces, organizations face a critical challenge: ensuring sensitive data isn't accidentally exposed. Traditional search mechanisms are no longer enough. To leverage the flexibility and power of LLMs without compromising security, you need a robust authorization strategy.

Join Oso Developer Advocate Greg Sarjeant, and Timescale AI Developer Advocate Jacky Liang, as they demonstrate how to build a secure, authorized LLM chatbot using Oso and Timescale. They’ll walk you through how to harness retrieval-augmented generation (RAG) and filter vector embeddings to ensure your chatbot only provides authorized responses.
‍

They'll cover:

• Why LLMs introduce new security challenges- and how unauthorized data exposure can occur.
• How to implement retrieval-augmented generation (RAG) to enhance chatbot accuracy and control.
• Best practices for filtering vector embeddings to enforce authorization rules.
• Integrating Oso and Timescale for seamless, scalable authorization.
• Practical tips for building a secure, production-ready chatbot.

If you have questions, stick around for the Q&A.

‍RSVP below - hope to see you there!