Europe 2026
Self-Training Agents: Hermes Agent, HF Traces, Skills, MCP & Finetuning — Merve Noyan, Hugging Face
Overview
This talk explores the burgeoning open-source AI agent ecosystem, emphasizing the accessibility and power of open models and tools. It highlights how advancements in Hugging Face's platform, including the Hub, Traces, Skills, and MCP tools, are democratizing the development and deployment of sophisticated AI agents. The core thesis is that open-source AI is rapidly catching up to and even surpassing closed-source alternatives, offering greater control, privacy, and customization for developers.
Who should watch
- AI Engineers looking to leverage open-source models and tools for agent development.
- Product Managers seeking to understand the capabilities and potential of agentic AI.
- Builders interested in prototyping and shipping AI-powered applications with greater autonomy.
- Developers concerned with data privacy and on-device deployment of AI.
- Anyone curious about the latest trends in agentic AI, including tool use and fine-tuning.
Key takeaways
- Open-source models are now competitive with, and in some cases surpass, closed-source models in performance, offering advantages like fine-tuning and local deployment for enhanced privacy.
- Hugging Face Hub serves as a central infrastructure for open-source AI workflows, hosting millions of models, datasets, and applications, with specific filters for agentic models.
- The Hugging Face ecosystem provides tools like MCP (Multi-modal Conversational Processing) for integrating LLMs with various services and Skills for enabling agents to perform complex tasks like model training and demo building.
- Hermes Agent is highlighted as a powerful tool for building AI agents, simplifying integration with platforms like Slack and WhatsApp, and can be effectively paired with strong open models like GLM 5.1.
- HF Traces is a new dataset repository type on Hugging Face Hub that allows for the hosting and exploration of agent execution traces, facilitating model training and analysis.
- Local deployment of AI models is increasingly accessible through tools like Llama CPP, VLLM, and MLX, with formats like GGUF enabling compatibility across various applications and hardware.
- Hugging Face Skills, such as the CLI skill, LLM trainer skill, and Gradio skill, empower coding agents to manage repositories, launch training jobs, and build interactive demos, streamlining complex AI development workflows.
- The MCP server allows agents to interact with Hugging Face Spaces, effectively acting as an App Store for AI, enabling functionalities like image generation through remote model calls.
Notable quotes
*Open source is absolutely differential.*
*If you have everything in the open, nothing changes without you knowing.*
*This is like a sci-fi at this point because it used to not exist.*
Unofficial community note. Prefer the recording for nuance.