AI Native Engineering

The presentation explores the paradigm shift towards enhancing productivity and creativity in software engineering through AI integration. The talk examines how AI can transform routine engineering tasks, allowing engineers to focus more on innovation and less on monotonous activities.

• Vision: Transitioning engineers from builders to explorers and innovators.
• Key Initiatives:
  • Reducing mundane tasks such as code review and testing.
  • Leveraging AI4P (AI for Productivity) program at Meta to improve engineering workflows.
• Outcomes:
  • Significant tool adoption and time savings in certain workflows.
  • Growth of a productivity-focused community from zero to over 400 members in seven months.
• Strategy:
  • Building a safe and encouraging space for experimentation with AI tools.
  • Implementing engineering excellence goals: quality, productivity, operational efficacy.
  • Creating patterns and documentation to aid in successful adoption.
• Challenges and Solutions:
  • Adopting new tools requires managing missteps and encouraging vulnerability.
  • Focus on integration quality and leveraging platform-specific context for enhanced tool effectiveness.
• Future Direction: Continuing to foster a community-driven approach for sustained improvement and adoption of AI in engineering.

Conclusion: The journey towards AI Native Engineering involves strategic planning, community building, and focused integration, all aimed at minimizing routine toil and maximizing creative engineering efforts.

This is the end of the AI-generated content.

Meta is making a big, public bet on AI - and not just in our products. Teams across the company are building new tools to leverage best-in-class models to enhance productivity, quality and understanding. As part of Horizon Experiences in Reality Labs, we have created a dedicated AI4P (AI For Productivity) programme to help teams effectively navigate the rapidly changing AI enhanced engineering  landscape. Our ambition is to help fulfil the vision of AI Native Engineering and this talk discusses what we’ve tried, what worked, what’s still to be proven and what we’ve discounted (at least for now).

We’ll discuss how we measure effectiveness of tools for our teams, how agentic AI has become part of our day to day coding activities and how standalone agents are working more like early career engineers than dumb machines. We’ll also talk about the leadership strategies employed to help people embrace and grow their understanding of the tools and associated techniques for optimal usage. Finally, we’ll look at how various platform integrations have led to step changes in output quality and engineer satisfaction.

Key Takeways

• Learn how product teams in Reality labs are leveraging AI tooling to support and enhance the full Software Engineering Lifecycle
• Understand how Engineering Excellence programmes hold the bar for quality, productivity and operational excellence, and the role that AI plays in raising that bar higher
• Hear about the strategies and tactics teams are using to create a culture of AI Native Engineering and embrace this new era of software development