At ELC 2025, what began as a discussion on defining outcome-oriented metrics for AI features turned into something more revealing: a shared struggle among engineering leaders to measure the real impact of AI-assisted work. This piece explores why teams often mistake productivity spikes for progress, the missing link between activity and outcome, and how to bridge that gap with metrics that actually matter.

James Reggio (CTO @ Brex) shares the story of "Brex 3.0", an 18-month journey behind their operational evolution. We explore how they rewound their org from a Series E to a Series C mindset, and replaced siloed OKRs with seasonal "marquee initiatives." James deconstructs the “Brex Hacker House”, an AI-focused startup within a startup experiment aimed to disrupt their core business. This conversation is all about evolving operational rhythms, layers of management, product building, and culture change!

AI is transforming how software gets built — from code generation to testing, release, and operations. But is it really paying off? In this session hosted by ELC and Harness, engineering leaders Adeeb Valiulla (Director, Technical Program Management at Harness) and Pushkar Priyadarshi (Director, Engineering at Harness) break down the data behind AI’s impact on software delivery. Learn why 63% of teams report faster development but nearly half experience more deployment problems — and how to move beyond speed to measurable outcomes. 💡 What You’ll Learn: The AI Velocity Paradox — why faster coding can slow overall delivery The three pillars of measurement: Utilization, Impact, and Cost How to avoid vanity metrics and track real business outcomes The role of guardrails and governance in sustaining AI ROI How leading teams use Harness SEI to prove the Return on Intelligence 🧠 Featuring Insights From: The State of AI in Software Engineering 2025 (Harness Report) Google DORA: State of AI-Assisted Software Development 2025 👉 Download the full reports: Harness 2025 Report (https://www.harness.io/the-state-of-ai-in-software-engineering) Google DORA 2025 Report (https://dora.dev/research/2025/dora-report/)

Businesses are spending millions on AI tools hoping to accelerate time-to-market but aren't seeing organizational-level results. Laura Tacho (CTO @ DX) explains why an "individual productivity" mindset fails and how AI merely accelerates the condition of the system it enters. She provides a framework for leaders to shift to a systems-level approach, find high-leverage ROI by looking outside the 20% of time spent coding, and understand what sets high-ROI orgs apart. Plus Laura shares data literacy tools to cut through the "whiplash" of conflicting AI reports and provides key considerations for 2026 budgeting, detailing where and how companies are planning to strategically invest.

What happens to platform engineering when natural language becomes the primary interface to infrastructure? Miriam Aguirre (Co-founder & CEO @ Ingenimax) joins us to explore how AI is fundamentally reshaping platform strategy, team structures, and the very role of the platform engineer. We deconstruct the shift from tactical "how" to strategic "why" and explore what it means to lead and build resilient systems in this new paradigm.