Now that everything is agentic, what should I be learning?
Summary
As traditional app logic shifts to agents, Microsoft recommends learning observability, evaluations, and evaluators. Filisha Shah explains that agents are non-deterministic, requiring a new approach to maintenance. Microsoft's AI Foundry platform integrates datasets, evaluations, and observability to manage these AI systems effectively.
Why it matters
This shift impacts developers building applications with AI agents. Traditional debugging and monitoring methods are insufficient for non-deterministic agents. Microsoft's AI Foundry aims to provide a unified solution for managing agent health and performance. This contrasts with competitors who might offer separate tools for these functions. Developers should monitor how platforms like AI Foundry evolve to handle the complexities of agent-based systems, especially as more logic moves to these autonomous components.
Related: OpenAI: First impressions of GPT-6 Astra from developers · Google: Graph Engineering 101 · Anthropic: Anthropic Cookbook: feat(claude_agent_sdk): add scheduled repository reviewer recipe (#860) · Meta: Nvidia launches free tool that links idle computers into a personal AI data center · xAI: Ajeya Cotra – "This might be the clearest warning shot we ever get"
Rated low: routine. Worth knowing, not worth rearranging your day for.
- Inside Microsoft's marketing team: scaling expertise with AIMicrosoft Azure on YouTube
- What does it really take to ship an AI agent?Microsoft Developer on YouTube
From Code to Agents: Build Production MCP Servers on Azure Functions
Questions people ask
- What should developers learn as app logic moves to agents?
- Developers should focus on learning observability, evaluations, and evaluators. This is because agents are non-deterministic and require a different approach to maintenance.
- What does Microsoft's AI Foundry offer?
- AI Foundry brings datasets, evaluations, and observability together in a single platform to help manage AI systems.
More from Microsoft Developer on YouTube 14 more
5 things you didn't know about Azure SQL Database Hyperscale | Data Exposed
Build Your Personal Brand with GitHub Copilot — Full Course for Students
From Code to Agents: Build Production MCP Servers on Azure Functions
Shield sensitive fields with Dynamic Data Masking for Microsoft SQL | Data Exposed MVP Edition
Bring Your Own Model in VS, Foundry Models, .NET Conf Sets Date | The Upload Ep 4
Page generated Sep 3, 2026. Summaries are Subvolts' own; the story belongs to Microsoft Developer on YouTube.