Best Practices for Agent Harness Based on MiniMax M2.7

Published on 14 May 2026, 16:53

Speaker: Ryan Lee, Lead Engineer, Developer Relationships, MiniMax
Talk Abstract: This session explores how MiniMax M2.7 enables self-evolving AI systems through scalable agent harnesses and high-throughput infrastructure. The talk covers the stack above the LLM — orchestration, memory, and tool use — alongside the infrastructure below the LLM, including optimized inference and distributed GPU systems, enabling a single developer to effectively consume and operationalize over one billion tokens per day.

Find the resources you need to develop using AMD products: amd.com/en/developer.html

Join the Developer Community: devcommunity.amd.com

Join the Developer Discord server: discord.gg/amd-dev

***

© 2026 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD Arrow logo, EPYC, ROCm, and AMD Instinct and combinations thereof are trademarks of Advanced Micro Devices, Inc.