GitHub for Leaders: Applying AI agents to software development

How Bayer boosted developer productivity and innovation (and job satisfaction) with AI.

In this episode of GitHub for Leaders, host Anjuan Simmons speaks with Mark Sparks, Senior Vice President and Head of Technology and Engineering at Bayer, to discuss practical strategies for integrating AI agents into the enterprise software development process—and how leaders can start doing the same right now.

What Bayer learned from adopting GitHub Copilot [00:46–2:01]

The conversation opens with Mark sharing Bayer’s data-driven approach to adopting GitHub Copilot. With results from their initial trial—including 89% of engineers reporting increased job satisfaction and 84% experiencing improved code quality—Mark underscores the impact of AI tools on engineering teams. 97% of Bayer engineers recommended Copilot to others.

Scaling AI across the enterprise [2:59–4:15]

Hear how Bayer integrated GitHub Copilot across diverse teams—from crop science to pharmaceuticals—creating a cross-functional and collaborative environment throughout the company. This approach empowered the developers to collaborate more effectively with key stakeholders spread throughout the product development lifecycle.

Rethinking software development with AI [4:25–6:18]

Mark also provides insights into how AI agents are reducing cognitive load by handling the manual labor of many complex tasks, and allowing engineers to focus on problem-solving rather than routine coding. He emphasizes the crucial shift from merely generating more code to generating the right code, and even deleting unnecessary code to streamline that process.

Training and mindset shifts for long-term impact [6:18–7:55]

Next, the conversation shifts to practical steps for implementing an AI-driven software development lifecycle—from forming cross-functional cohorts for initial adoption, to critical mindset shifts through targeted training. Mark also shares strategies to unlock productivity gains for both junior engineers and seasoned veterans by leveraging their experience to guide AI-driven outputs more effectively.

Building an AI-native development cycle [7:55–8:54]

As the discussion wraps up, the conversation explores the evolving capabilities of AI agents, from leveraging large language models for advanced reasoning and planning to enabling collaborative workflows across multiple agentic functions within an organization.

👉 Whether you’re considering adopting AI agents or scaling existing implementations, this episode provides actionable guidance, real-world insights, and a clear path forward for tech leaders aiming to harness the full potential of AI.

Want to learn more about the strategic role of security and other innovations at GitHub? Explore Executive Insights for more thought leadership on the future of technology and business.