The Renaissance Developer: How Agentic AI is redefining software engineering

For many years developer productivity improved through better frameworks, faster runtimes, and stronger cloud platforms. Today, the biggest shift is happening elsewhere: how we work.

In this session, I’ll demonstrate how agentic AI tools and LLMs are transforming everyday development tasks: from planning the project to coding, debugging and code reviewing, documentation and data extraction.

We’ll explore how traditional techniques are being replaced or augmented by LLM-based approaches, such as using natural-language reasoning instead of complex parsing logic based on regular expressions. The session includes practical examples using tools like Kiro, Claude Code, and foundation models hosted on Amazon Bedrock.

Beyond the tools themselves, I’ll show how developers can extend and guide AI assistants: teaching them new skills, integrating external knowledge through MCPs, managing context effectively and aligning them with existing development workflows and coding standards. These techniques allow AI agents to operate confidently even with cutting-edge AWS services and newly released technologies that were not available when the LLM’s were trained.

If you want a realistic look at what modern development is becoming, this session will give you a glimpse into the next era of software engineering.