Description

In this episode of the Data & AI Podcast, host Alex Juarez, Director of Engineering at Mesh-AI, is joined by Elliot Budd (Lead AI Engineer), Millie Wan Marriott (Data Engineer), and Maartens Lourens (Principal AI Engineer) for a deep dive into the rapidly evolving world of AI-assisted coding.

The team explores how AI coding tools have transformed from basic autocomplete suggestions to sophisticated development partners - and what that means for the future of software engineering. From the early days of GitHub Copilot to today's agentic coding systems, they discuss the trust, limitations, and best practices that are shaping how developers work alongside AI.

This episode covers:

• The evolution from autocomplete to agentic coding systems
• Why trust is the critical factor in adopting AI coding tools
• The gap between "vibe coding" and proper engineering practices
• Current limitations: context management, hallucinations, and when AI changes the wrong code
• Emerging best practices and the concept of "guardrail engineering"
• What junior developers need to learn in an AI-assisted world
• The future of verification, testing, and developer satisfaction

Whether you're sceptical about AI coding tools or already using them daily, this conversation offers honest insights into what works, what doesn't, and where the industry is headed.

"Developer satisfaction is at an all-time high with these tools. If five years ago it took you a week to build an API, and now you can do it in an hour - that's really satisfying."

Hosted on Acast. See acast.com/privacy for more information.