Leveraging AI Agents for Architecture and Project Planning Work

AI coding agents are no longer just developer assistants—they’re transforming how architects and project leaders plan and execute change. This session explores how agentic AI, such as Claude Code agents, can analyze existing systems, identify opportunities for modernization, and streamline project planning for large and complex codebases.

Through real-world examples, attendees will see how AI agents accelerate architectural reviews, dependency analysis, and multi-phase enhancement planning. You’ll learn practical methods for integrating agentic AI into your architecture and planning workflows to reduce manual effort and improve decision-making.

While examples center on Claude Code, the insights apply broadly as this rapidly evolving field continues to develop. Attendees will leave knowing how to utilize AI agents to streamline architecture and planning processes, working faster, smarter, and more strategically.

AI-Powered Application Modernization: From Legacy Code to Cloud-Native with Claude-Flow

Application modernization projects often stall under the weight of complexity: Do you refactor, re-platform, or replace with SaaS? AI-assisted development can cut through that complexity, turning weeks of manual analysis into hours.

In this session, I’ll share a real-world exercise where my Claude-Flow swarm of coding agents was used extensively to modernize an existing open-source CRM application. We’ll cover three stages of the journey: (1) assessing the application’s fitness for cloud deployment, (2) using AI to analyze the viability of migrating users to existing SaaS alternatives, and (3) planning the modernization effort with a roadmap that accounts for technical debt, risk, and compliance gaps.

This is not a theoretical discussion—I’ll show concrete outputs generated by the swarm, how the orchestration handled complex decomposition, and where human judgment was still required. Attendees will walk away with a practical, field-tested framework for applying AI agents to modernization efforts and accelerating the move to cloud-native architectures.

Scaling AI Assistance: Why Agent Teams Change Development

As AI-assisted development matures, I’m seeing single-agent coding workflows hit practical limits around context, parallelism, and task focus. In this session, I explore when and why coordinated teams of AI agents outperform single-agent approaches—and when they do not.

I present a technical comparison of Claude Code Agent Teams, Claude-Flow, and other mainstream agent orchestrators, examining how different coordination models handle task decomposition, shared context, write boundaries, and control flow. Through concrete examples, I show how these design choices affect correctness, throughput, and developer oversight.

I demonstrate how agent teams reshape development across the software lifecycle, from requirements exploration and architecture design to implementation, testing, and operational support. I contrast single-agent and multi-agent workflows to highlight where specialization and parallel reasoning improve outcomes, and where simpler models remain preferable.

Attendees will leave with a clear framework for deciding when to scale AI assistance using agent teams, how to evaluate orchestration tooling, and how to apply these techniques effectively in real-world engineering environments.

Given how quickly this space changes, new products or other recent advances may be mentioned closer to presentation day.

AI-Powered Application Modernization: From Legacy Code to Cloud-Native with Agentic AI

Application modernization projects often stall under the weight of complexity. Teams are forced to choose between refactoring, re-platforming, or replacing systems with SaaS, often with incomplete information and high risk. Agentic AI and AI-assisted development approaches can cut through this complexity by transforming weeks of manual analysis into hours of structured, decision-ready output.

In this session, I’ll walk through a real-world modernization exercise in which coordinated teams of AI coding agents analyzed and planned the modernization of an existing open-source CRM application. The session begins by using AI agents to assess the application’s overall fitness for cloud deployment, including architectural constraints, operational risks, and technical debt that could affect cloud readiness.

From there, the talk explores how agent teams can be applied to evaluate whether modernization is even the right path. AI agents are used to research and compare existing SaaS alternatives, analyze migration feasibility, and surface trade-offs around functionality, data migration, cost, and organizational impact—providing leadership with clearer options earlier in the decision process.

Finally, I’ll show how agentic AI can be used to produce a practical modernization roadmap. This includes identifying modernization phases, sequencing work to reduce risk, and accounting for compliance, security, and operational gaps. Throughout the session, I’ll highlight the artifacts produced by agent teams, how orchestration and task decomposition work in practice, and where human judgment and architectural decision-making remain essential.

This is not a theoretical discussion. Attendees will see concrete outputs from a real modernization effort and leave with a repeatable, tool-agnostic framework for applying agentic AI to application modernization initiatives and accelerating the move to cloud-native architectures.

While Claude Code Agent Teams and Claude-Flow are top of mind right now, other products might be included if the product landscape changes.

From Vibe Coding to Spec-Driven Development: Governing Intent at the Speed of AI

AI-assisted development has made it possible to generate working code from little more than a prompt—but as systems grow, “vibe coding” quickly breaks down. Requirements drift, architectural decisions are lost in chat history, and teams are left reverse-engineering intent from generated code. For engineers, architects, and team leads responsible for real systems, this creates risk, rework, and governance challenges.

This session introduces Spec-Driven Development (SDD) as a pragmatic evolution of AI-assisted engineering. Instead of treating prompts as ephemeral instructions, SDD formalizes intent into durable, versioned specifications that guide AI agents throughout design, implementation, and change. We’ll walk through the core SDD lifecycle—proposal, review, implementation, and archival—and show how it complements modern DevOps and architecture practices rather than replacing them.

Using the OpenSpec framework, we’ll explore how teams can decouple requirements from chat sessions, preserve architectural decisions, and give AI coding agents clear, testable boundaries. Attendees will see how specs become a shared contract between humans and AI, enabling faster iteration without sacrificing clarity, traceability, or control.

This talk focuses on practical patterns you can apply immediately to scale AI-assisted development responsibly across teams and systems.