AI Coding Summit

AI coding assistants are great at completing small tasks or features. However, what do you do when you are working with more complex code bases, and you need to build in-depth features that need upfront planning?This talk explores spec-driven development as a solution to this problem. I'll show you how modern AI coding assistants (like Kiro) can help break down complex tasks into three distinct phases. We'll look at the real-world tradeoffs of this approach, and most importantly and how you can use it in your own projects right away.

AI agents are no longer experimental. Developers are already using them to query APIs, modify content, and chain services using emerging protocols like MCP (Model Context Protocol). The latest MCP specification introduces modern OAuth 2.1 authentication and support for Resource Indicators (RFC 8707), strengthening identity in agent-based systems.

But authentication alone does not guarantee control. Once an agent is logged in, how do you govern what it is allowed to do? Without proper authorization controls, agents can access far more resources than they need, creating significant security risks.

This talk explores how to apply Zero Trust principles to agent workflows by combining open identity protocols with policy-aware infrastructure. You will see a demo of an MCP client interacting with a secured MCP server behind Pomerium, an open source identity-aware proxy that brings fine-grained access control to agent interactions. Beyond basic authentication, Pomerium evaluates per-request policies based on identity, route, and context, and can audit and block specific tool calls within the MCP protocol. It can even manage OAuth flows to upstream tools like Notion or Reddit, so agents never handle raw access tokens.

What you will learn:

- Why OAuth is necessary but not sufficient for agent security
- How to apply Zero Trust to developer tools and AI workflows
- A practical example of securing MCP servers with open source infrastructure

As AI agents become part of real-world developer workflows, open standards and secure defaults are key to building trust without adding friction. These security patterns apply beyond just AI systems to any automated tooling that needs controlled access to APIs and services.

Over the past few years, there has been a notable shift in how developers interact with documentation.Before now, developers often turned to Docs and Stack Overflow for answers about tools, but that is no longer the case.Today, API Docs, tutorials, and specifications are not just for humans—they are also utilized to train LLM agents to answer questions about your tools. This talk will explore how to document API and tutorials in a way that serves both developers and LLMs, making your documentation more accessible, actionable, and future-proof.Through case studies and practical insights, we'll cover how companies are adjusting to this new dynamic shift and the rise of GEO (Generative Engine Optimization).

This talk shows how to wire agentic workflows into modern CI/CD so code gets safer and faster to prod: LLM code review as a first pass, automated remediation suggestions, and swarm-style orchestration for follow-ups. You’ll see what actually works today with Copilot Code Review, CodeQL Autofix, Semgrep Assistant, and agent frameworks like LangGraph or AutoGen, plus how to gate merges with lightweight, automated evaluations.

We've all heard about prompt engineering. But now with the emergence of context engineering you may be scratching your head about what the difference is. The reality is that to build reliable agents, we need to use all tools in our context toolbox to ensure they provide accurate responses in the correct format.Let's dive into the world of context engineering, and understand the elements of context that we can tune and tweak to build more reliable AI agents and associated systems.

As AI moves into the heart of software development, the role of developers is shifting from writing code to shaping systems. This talk explores how the classic principle of Single Responsibility evolves in an AI-first world, why it’s the key to scalable design, and how it transforms every developer into an architect.

What happens when generative AI isn't just a buzzword but a real engineering teammate? Over the past year, I've led my engineering team through a radical shift-integrating GenAI tools and building our own custom AI-powered solutions. The result? We've multiplied productivity, cut delivery times, and unlocked new creative workflows that were impossible before.In this talk, I'll share real-world examples of how we use tools like Cursor and Devin, as well as custom-built AI integrations, to speed up coding, automate testing, improve documentation, and even resolve complex bugs in minutes. I'll break down the process changes that made AI adoption successful-how we rewired our workflows, onboarded the team, and learned to "pair program" with AI effectively. Expect practical lessons and battle-tested insights: what worked brilliantly, what failed (and why), and the cultural shifts needed to make GenAI a force multiplier instead of just another shiny tool. Whether you're a team lead, engineer, or product builder, you'll walk away with actionable ideas to supercharge your own team using GenAI-without losing the human touch that drives great engineering.

Discover how to use Docker to package, ship, and run AI tools built on the Model Context Protocol (MCP). Learn how to create reproducible MCP servers and clients, integrate them into developer workflows, and scale them in production — from local dev to agentic CI/CD pipelines.

AI-powered code review promises faster development cycles, but making it work in production requires more than just plugging in an LLM. This talk explores best practices, pitfalls, and real-world lessons from building AI-driven review and refactoring pipelines in enterprise systems. Attendees will learn how to evaluate quality, avoid false positives, and integrate AI review safely into CI/CD workflows.

What if you let AI code alongside your devs—would it speed things up or slow them down? At Modus Create, we ran a 3-month internal experiment to find out. Two squads were tasked with building the same app: one squad had full access to AI coding agents like GitHub Copilot and Cursor, while the other worked with zero AI assistance. The results? Surprising, nuanced, and extremely relevant for any enterprise thinking about bringing AI tools into their software delivery process. In this session, we'll share productivity metrics, insights on code quality, cultural impact, and what we learned about pairing developers with AI. You'll walk away with a real-world perspective and a practical guide on how to scale AI-powered development across your engineering org—without the hype.

In this session, I’ll explore how AI-powered tooling is transforming large-scale refactoring and codebase migrations, making these complex tasks faster and more efficient. By leveraging tools like Large Language Models (LLMs), static analysis, and refactoring frameworks, we can automate repetitive code transformations, accelerate migration paths, and reduce human error. I’ll share practical examples of how we can migrate legacy systems to modern frameworks, break monolithic architectures into service-based structures, and automate large-scale code changes across thousands of files. I’ll demonstrate how tools like OpenAI Codex, GitHub Copilot, and fine-tuned models for domain-specific transformations can assist in these processes, while still integrating with traditional migration tools and CI/CD systems. I’ll also cover the importance of developer oversight, highlighting lessons learned from real-world production rollouts and how to balance automation with manual reviews.By the end of the talk, I want attendees to walk away with a clear understanding of how to use AI to enhance their refactoring and migration workflows, ensuring they can handle large-scale transformations while maintaining code quality, consistency, and governance.