Working with Models¶
Explore labs, session repositories, and additional resources from Microsoft Build 2026 focused on working with AI models, including training, fine-tuning, and deploying models for real-world applications.
Session Repositories¶
| Session | Presenters | Repo | Session Page |
|---|---|---|---|
| Build smarter AI systems in Foundry as models and costs evolve | Yina Arenas, Naomi Moneypenny | Repo | BRK230 |
| Deploy. Observe. Learn. Reinforcement learning for production agents | Omkar More, Alicia Frame | Repo | BRK231 |
| Post-Training and Deploying Open Source Reasoning Models in Foundry | Vijay Aski, Chris Lauren, Manoj Bableshwar | Repo | BRK232 |
| Hugging Face open‑source models to production on Microsoft Foundry | Vaidya Sambasivam, Osi Otugo, Jeff Boudier | Repo | DEM320 |
| Get Started with Models in Microsoft Foundry to Build AI Apps | Lee Stott, Aaron Powell | Repo | LAB520 |
| Improving agent behavior using reinforcement learning from traces | Alicia Frame | Repo | LAB521 |
📚 Additional Resources¶
Model Selection and Foundry¶
- Microsoft Foundry documentation - Overview of foundation models and model deployment in Microsoft Foundry.
- What is Microsoft Foundry? - Platform capabilities for training, deploying, and managing AI models.
- Train and deploy models with Microsoft Foundry - Hands-on learning path for ML workflows in Azure.
Fine-Tuning and Custom Models¶
- Fine-tune language models in Microsoft Foundry - Practical guidance for adapting foundation models to specialized tasks.
- Responsible AI best practices for fine-tuning - Bias mitigation and responsible model adaptation strategies.
- Model distillation and optimization - Techniques for creating smaller, faster models for edge and inference.
Reinforcement Learning and Agent Training¶
- Reinforcement learning concepts and patterns - Core RL theory and practical agent design patterns.
- Training agents with Microsoft Foundry - End-to-end RL training pipeline setup and orchestration.
- Evaluating and improving agent behavior - Monitoring, tracing, and continuous improvement for production agents.
MLOps and Model Deployment¶
- MLOps fundamentals - Version control, automation, and lifecycle management for ML models.
- Model registry and deployment - Governance and production deployment strategies.
- Monitoring and observability for deployed models - Drift detection, performance tracking, and alerting for live models.