2026-06-18 –, E105 (capacity 70)
Most teams use AI in silos — quietly, individually, with an unspoken awkwardness about what's AI-generated. The result: faster output with hidden fragility.
The first shift was making AI official: a versioned context file, shared across the team, updated with every feature. AI stopped being stateless. Early results showed 40–50% fewer prompt iterations.
But shared context kept failing when engineers weren't clear on the problem before prompting. AI output quality tracks directly to problem clarity — and that was the missing piece. So we built IDEA: Intent, Diagnose, Explain, Architect — a mental model for thinking before prompting.
We'll demo IDEA live, show a real context file, and share what broke along the way. You'll leave with the IDEA framework, a context file template, and a prompt quality checklist.
I am a Software Engineering Manager at Red Hat with over 12 years of IT experience. I lead teams in designing, developing, and delivering high-quality, scalable software. My role includes promoting engineering best practices, establishing quality processes, and ensuring products meet technical and business goals. I focus on continuous improvement, fostering collaboration, and working with cross-functional teams to deliver reliable, impactful products.
Pranjal Bathia is a Senior Engineering Manager at Red Hat, leading the transformation of its MDM platform into an AI-native system. With 10+ years’ experience, she drives product, pricing, and integration modernization, building scalable systems and AI-driven workflows. She actively promotes AI adoption and fosters strong teams, while enjoying creative time with her daughters outside work.