Evolve to AI

At monday.com, we set out to take coding agents beyond experimentation, with a clear mission: rethink the SDLC end-to-end using purpose-built autonomous agents. By offloading repetitive and operational work to dev agents, our engineers can stay focused on the problems that truly require human judgment.

In this talk, I’ll share the vision, technical/product challenges, and hard-won lessons behind building monday’s internal Continuous AI platform - a system that enables any builder to create agentic workflows with ease. Powered by Claude Code, Claude Agent SDK, and GitHub workflows, the platform democratizes agent building across the organization while keeping it practical, scalable, and production-ready.

The future of AI isn't just about better models; it's about better interfaces. How do we enable agents to interact with users beyond simple text streams? In this session, we will unpack the MCP UI and A2UI stack as the foundation for modern agentic workflows. We will show how to build an architecture where agents can dynamically orchestrate their own UI to visualize thinking, request permissions, and display results. You will walk away with practical techniques for upgrading your agent interactions from "black box" execution to transparent, rich visual collaboration.

Most AI agents today suffer from digital amnesia: they forget conversations, lose context between sessions, and treat every interaction as if meeting you for the first time. We explore how knowledge graphs can serve as persistent, temporally-aware memory systems for AI agents, moving beyond static vector embeddings to dynamic, relationship-rich storage that evolves with each interaction. This talk covers practical implementation patterns for building graph-based agent memory, comparing direct database integration with tool-based approaches through frameworks like Graphiti and Zep. You'll learn how to extract entities from conversations, model temporal relationships, and implement cross-session continuity that makes agents truly conversational partners rather than stateless responders.

What if we could move a big chunk of bug fixing and solving production issues to agentic AI? That would be so cool. In this talk we will go through the end to end process of setting up a background agentic workflow that detects production errors, finds their root causes, assesses the right solution and opens a PR - so you wake up in the morning to tasks almost fully completed for you by your loyal agent.

Together We will go over the entire process - the infra challenges, the agentic accuracy issues - and the final result used nowadays in real, massive production environments.

Production issues still slow teams down. Finding the root cause is noisy, fixing it is manual, and the cycle continues. But how do you break that loop? In this session, you’ll see how you can wire production context into an AI agent that not only detects failing endpoints, but also traces the root cause, proposes a fix, and opens a pull request - automatically.

Join us to see how AI can take the sting out of bug triage, accelerate recovery, and boost the happiness of both your team and your users.

We are in a rare "capability overhang" moment, a window where AI models possess far more potential than current applications actually leverage. This session shares Microsoft's practical insights and architectural best practices from real-world startups and applications on moving beyond the chatbot toward autonomous systems. We will explore how to address the gap between "cool demo" and a "production-grade agent" and build agentic workflows that solve complex, multi-step problems in production.

Focusing on the "under-the-hood" mechanics, we’ll dive into the transition from reactive assistants to goal-oriented acritarches using proven patterns like ReAct, Reflexion, and Planner-Executor. During the session, we will deep dive into best practices on navigating the technical trade-offs of modern frameworks such as Microsoft Agent Framework, AutoGen, and LangGraph to address key components of real-world applications such as state management, persistent memory, and cost optimization.