Build time: RFT Agent
Build time: RFT Agent
Author: OpenAI – Duration: 01:02:11
Agent RFT enables reasoning models to become agents using even more powerful tools by training directly on the workflows they will run in production. Operating on agent deployment, reasoning models can invoke tools, generate intermediate reasoning steps, and receive real-time feedback through customer-provided endpoints. This build hour will review the preparation, infrastructure, and security oversight for using Agentic RFT. Theophile Sautory (Applied AI) and William Hang (API Engineering) cover: • Improving agent performance with optimization and fine-tuning options. • Main differences between Base RFT and Agentic RFT. • New additions and operation of the RFT agent. • Task configuration and live demonstrations with tools. • Spotlight on Cognition with Sampriti Panda (research engineer). https://tinyurl.com/agentRFT
👉 Follow the code repository: https://github.com/openai/build-hours
👉 Sign up for upcoming live build times: https://webinar.openai.com/buildhours/
00:00 Introduction 01:34 Introduction to Agent RFT 11:12 Configuring Tasks 14:15 Demos: Training with the Tools 31:33 Best Practices 35:15 Customer Spotlight: Cognition 44:58 Success Stories 51:16 Summary 52:33 Questions and Answers
