Update AI Engineer Summit talk pages with video timestamps and standardized titles

- Reorganized video timestamps in schedule page to nest under talk page links
- Updated all talk page titles to include 'Talk:' format matching Anthropic example
- Added video timestamps to Talk section headers for all talks with available timestamps

Author: codekiln

pages/AI___ES___25___11 Code___LEAD___20 Thu.md

@@ -15,29 +15,41 @@
 	- [[AI/ES/25/11 Code/LEAD/20 Thu/1445 Arc to Dia Lessons from Browser Company Samir Mody]]
 - ## Video timestamps
 	- [00:03:02](https://www.youtube.com/watch?v=cMSprbJ95jg&t=182s) - "The New Code" and the Philosophy of Creation
-	  [00:13:38](https://www.youtube.com/watch?v=cMSprbJ95jg&t=818s) - Host Alex Lieberman Welcomes Attendees & Introduces the Summit
-	  [00:17:45](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1065s) - Anthropic (Katelyn Lesse): 3 Pillars of Building Agentic Systems with Claude
-	  [00:31:36](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1896s) - Replit (Michele Catasta): Supervised vs. Unsupervised Autonomy for Non-Technical Users
-	  [00:57:46](https://www.youtube.com/watch?v=cMSprbJ95jg&t=3466s) - Zapier (Lisa Orr): Case Study on "Scout," the Support Team AI Agent
-	  [01:10:31](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4231s) - Steve Yegge & Gene Kim Begin Their High-Energy "Vibe Coding" Talk
-	  [01:16:54](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4614s) - Hot Take: Steve Yegge Declares the IDE is Dead and Using One Makes You a "Bad Engineer"
-	  [01:22:22](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4942s&pp=0gcJCTAAlc8ueATH) - Gene Kim Defines "Vibe Coding" & the FAFO Framework (Faster, Ambitious, Fun, Optionality)
-	  [01:35:20](https://www.youtube.com/watch?v=cMSprbJ95jg&t=5720s) - Mid Morning Break Begins
-	  [02:11:38](https://www.youtube.com/watch?v=cMSprbJ95jg&t=7898s) - OpenAI Team (Bill Chen & Brian Fioca) Presentation Starts
-	  [02:15:16](https://www.youtube.com/watch?v=cMSprbJ95jg&t=8116s) - OpenAI Explains the Importance of the Agent "Harness"
-	  [02:35:51](https://www.youtube.com/watch?v=cMSprbJ95jg&t=9351s) - McKinsey (Martin Harrysson & Natasha Maniar): Argument that the Agile Operating Model is a Bottleneck
-	  [02:55:05](https://www.youtube.com/watch?v=cMSprbJ95jg&t=10505s) - Stanford (Yegor Denisov-Blanch): Research Shows Codebase Cleanliness Correlates Strongly with AI Productivity
-	  [03:12:12](https://www.youtube.com/watch?v=cMSprbJ95jg&t=11532s) - Qodo (Itamar Friedman): Breaking the "Glass Ceiling" of Productivity with Agentic Quality
-	  [03:42:30](https://www.youtube.com/watch?v=cMSprbJ95jg&t=13350s) - Lunch Break Begins
-	  [05:00:11](https://www.youtube.com/watch?v=cMSprbJ95jg&t=18011s) - Google Labs (Kath Korevec): The Shift from Reactive to Proactive Agents ("Jules")
-	  [05:22:33](https://www.youtube.com/watch?v=cMSprbJ95jg&t=19353s) - Northwestern Mutual (Asaf Bord): Case Study on a Phased Rollout of a GenBI Agent
-	  [05:46:16](https://www.youtube.com/watch?v=cMSprbJ95jg&t=20776s) - Bloomberg (Lei Zhang): Creating a "Paved Path" for Enterprise AI Adoption
-	  [06:03:13](https://www.youtube.com/watch?v=cMSprbJ95jg&t=21793s) - The Browser Co. (Samir Mody): Treating "Model Behavior" as a Craft & Discipline
-	  [06:29:07](https://www.youtube.com/watch?v=cMSprbJ95jg&t=23347s) - Capital One (Max Kanat-Alexander): Core Principle: "What's good for humans is good for AI"
-	  [07:09:56](https://www.youtube.com/watch?v=cMSprbJ95jg&t=25796s) - Final Block of Sessions Begins
-	  [07:19:56](https://www.youtube.com/watch?v=cMSprbJ95jg&t=26396s) - NLW (AI Daily Brief): Initial Findings from Self-Reported AI ROI Study
-	  [07:29:07](https://www.youtube.com/watch?v=cMSprbJ95jg&t=26947s) - Hot Take: 10X (Arman Hezarkhani) Explains the Model for Paying Engineers Like Salespeople (Per Story Point)
-	  [08:10:27](https://www.youtube.com/watch?v=cMSprbJ95jg&t=29427s) - Every (Dan Shipper): "Compounding Engineering" in an AI-Native Company
-	  [08:19:22](https://www.youtube.com/watch?v=cMSprbJ95jg&t=29962s) - Host Alex Lieberman's Closing Remarks & After-Party Kickoff
+	- [00:13:38](https://www.youtube.com/watch?v=cMSprbJ95jg&t=818s) - Host Alex Lieberman Welcomes Attendees & Introduces the Summit
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/0905 Anthropic Claude APIs Agents]]
+		- [00:17:45](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1065s) - Anthropic (Katelyn Lesse): 3 Pillars of Building Agentic Systems with Claude
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/0925 Michele Catasta VP Replit Autonomy is all you need]]
+		- [00:31:36](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1896s) - Replit (Michele Catasta): Supervised vs. Unsupervised Autonomy for Non-Technical Users
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/0945 Lisa Orr Zapier Your Support Team Shipping Code]]
+		- [00:57:46](https://www.youtube.com/watch?v=cMSprbJ95jg&t=3466s) - Zapier (Lisa Orr): Case Study on "Scout," the Support Team AI Agent
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1005 2026 year ide died steve yegge gene kim amp sourcegraph]]
+		- [01:10:31](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4231s) - Steve Yegge & Gene Kim Begin Their High-Energy "Vibe Coding" Talk
+		- [01:16:54](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4614s) - Hot Take: Steve Yegge Declares the IDE is Dead and Using One Makes You a "Bad Engineer"
+		- [01:22:22](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4942s&pp=0gcJCTAAlc8ueATH) - Gene Kim Defines "Vibe Coding" & the FAFO Framework (Faster, Ambitious, Fun, Optionality)
+	- [01:35:20](https://www.youtube.com/watch?v=cMSprbJ95jg&t=5720s) - Mid Morning Break Begins
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1100 OpenAI Applied AI Bill Chen Brian Fioca Future-Proof Coding Agents]]
+		- [02:11:38](https://www.youtube.com/watch?v=cMSprbJ95jg&t=7898s) - OpenAI Team (Bill Chen & Brian Fioca) Presentation Starts
+		- [02:15:16](https://www.youtube.com/watch?v=cMSprbJ95jg&t=8116s) - OpenAI Explains the Importance of the Agent "Harness"
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1120 McKinsey Moving away from Agile]]
+		- [02:35:51](https://www.youtube.com/watch?v=cMSprbJ95jg&t=9351s) - McKinsey (Martin Harrysson & Natasha Maniar): Argument that the Agile Operating Model is a Bottleneck
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1140 Yegor Denisov-Blanch Stanford AI ROI Software Engineering]]
+		- [02:55:05](https://www.youtube.com/watch?v=cMSprbJ95jg&t=10505s) - Stanford (Yegor Denisov-Blanch): Research Shows Codebase Cleanliness Correlates Strongly with AI Productivity
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1200 Itamar Friedman CEO Qodo State of AI Code Quality]]
+		- [03:12:12](https://www.youtube.com/watch?v=cMSprbJ95jg&t=11532s) - Qodo (Itamar Friedman): Breaking the "Glass Ceiling" of Productivity with Agentic Quality
+	- [03:42:30](https://www.youtube.com/watch?v=cMSprbJ95jg&t=13350s) - Lunch Break Begins
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1345 Proactive Agents Kath Korevec Google Labs]]
+		- [05:00:11](https://www.youtube.com/watch?v=cMSprbJ95jg&t=18011s) - Google Labs (Kath Korevec): The Shift from Reactive to Proactive Agents ("Jules")
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1405 Northwestern Mutual Asaf Bord Small Bets Big Impact]]
+		- [05:22:33](https://www.youtube.com/watch?v=cMSprbJ95jg&t=19353s) - Northwestern Mutual (Asaf Bord): Case Study on a Phased Rollout of a GenBI Agent
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1425 Lei Zhang Bloomberg Tech Infra Eng]]
+		- [05:46:16](https://www.youtube.com/watch?v=cMSprbJ95jg&t=20776s) - Bloomberg (Lei Zhang): Creating a "Paved Path" for Enterprise AI Adoption
+	- [[AI/ES/25/11 Code/LEAD/20 Thu/1445 Arc to Dia Lessons from Browser Company Samir Mody]]
+		- [06:03:13](https://www.youtube.com/watch?v=cMSprbJ95jg&t=21793s) - The Browser Co. (Samir Mody): Treating "Model Behavior" as a Craft & Discipline
+	- [06:29:07](https://www.youtube.com/watch?v=cMSprbJ95jg&t=23347s) - Capital One (Max Kanat-Alexander): Core Principle: "What's good for humans is good for AI"
+	- [07:09:56](https://www.youtube.com/watch?v=cMSprbJ95jg&t=25796s) - Final Block of Sessions Begins
+	- [07:19:56](https://www.youtube.com/watch?v=cMSprbJ95jg&t=26396s) - NLW (AI Daily Brief): Initial Findings from Self-Reported AI ROI Study
+	- [07:29:07](https://www.youtube.com/watch?v=cMSprbJ95jg&t=26947s) - Hot Take: 10X (Arman Hezarkhani) Explains the Model for Paying Engineers Like Salespeople (Per Story Point)
+	- [08:10:27](https://www.youtube.com/watch?v=cMSprbJ95jg&t=29427s) - Every (Dan Shipper): "Compounding Engineering" in an AI-Native Company
+	- [08:19:22](https://www.youtube.com/watch?v=cMSprbJ95jg&t=29962s) - Host Alex Lieberman's Closing Remarks & After-Party Kickoff
 - ## Companies Mentioned
 	- [[Anthropic]], [[ReplitAI]], [[Zapier]], [[Sourcegraph]], [[OpenAI]], [[McKinsey]], [[QodoAI]], [[MiniMaxAI]], [[Google]], [[Northwestern Mutual]], [[Bloomberg]], [[Browser/Company]], [[Monday.com]], [[Nubank]], [[Salesforce]], [[Sonar]], [[Faros]], [[Nvidia]], [[Zed]], [[Booking.com]], [[Cisco]], [[Fidelity]], [[Travelopia]]

pages/AI___ES___25___11 Code___LEAD___20 Thu___0905 Anthropic Claude APIs Agents.md

@@ -1,7 +1,7 @@
-- # 9:05am - 9:24am | AI Leadership | Room: Times Center
+# 9:05am - 9:24am Talk: Evolving Claude APIs for Agents
 	- ![Katelyn Lesse](https://www.ai.engineer/speakers/katelyn-lesse.jpg)
 	- **[[Person/Katelyn Lesse]]** [Twitter](https://twitter.com/katelyn_lesse) [LinkedIn](https://www.linkedin.com/in/katelynlesse) - Head of Engineering, Claude Developer Platform, Anthropic
-	- ## Talk: Evolving Claude APIs for Agents
+	- ## Talk: Evolving Claude APIs for Agents [00:17:45](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1065s) - Anthropic (Katelyn Lesse): 3 Pillars of Building Agentic Systems with Claude
 		- Developers are building more and more complex, long-running, agentic systems. Learn how the Anthropic team is evolving the Claude Developer Platform to enable developers to get the best outcomes from Claude.
 	- ## Memory Tool
 		- Originally just [[Anthropic/App/Claude Code/Claude.md]] file based

pages/AI___ES___25___11 Code___LEAD___20 Thu___0925 Michele Catasta VP Replit Autonomy is all you need.md

@@ -1,7 +1,7 @@
-- # 9:25am - 9:44am | AI Leadership | Room: Times Center
+# 9:25am - 9:44am Talk: Autonomy Is All You Need
 	- ![Michele Catasta](https://www.ai.engineer/speakers/michele-catasta.jpg)
 	- **[[Person/Michele Catasta]]** [Twitter](https://twitter.com/pirroh) [LinkedIn](https://www.linkedin.com/in/pirroh) [GitHub](https://github.com/pirroh) - VP of AI, Replit
-	- ## Talk: Autonomy Is All You Need
+	- ## Talk: Autonomy Is All You Need [00:31:36](https://www.youtube.com/watch?v=cMSprbJ95jg&t=1896s) - Replit (Michele Catasta): Supervised vs. Unsupervised Autonomy for Non-Technical Users
 		- AI agents exhibit vastly different degrees of autonomy. Yet, the ability to accomplish objectives without supervision is the critical north star for agent progress, especially in software creation. For non-technical users who cannot supervise software creation, full autonomy is essential, not optional.
 		- First of all, I will discuss two foundational capabilities to achieve true autonomy: automatic testing to verify correctness without human validation, and advanced context management to maintain coherence across complex, long-horizon tasks.
 		- With autonomy established, parallelization becomes the key to delivering a compelling user experience. Sequential execution forces users to wait extensively before seeing progress, breaking the development flow. This talk explores parallelization models (task-level parallelism, out-of-order execution, plan decomposition, etc.) and their tradeoffs in latency, resource consumption, and correctness guarantees.

pages/AI___ES___25___11 Code___LEAD___20 Thu___0945 Lisa Orr Zapier Your Support Team Shipping Code.md

@@ -1,9 +1,9 @@
 chatgpt-link:: https://chatgpt.com/g/g-p-691f249c2fac8191ab8b4b926da5cb3b-ai-es-25-11-code/c/691f2a43-888c-800a-bd67-c3657e6931d4
 
-- # 9:45am - 10:04am | AI Leadership | Room: Times Center
+# 9:45am - 10:04am Talk: Your Support Team Should Ship Code
 	- ![Lisa Orr](https://www.ai.engineer/speakers/lisa-orr.jpg)
 	- **[[Person/Lisa Orr]]** [Twitter](https://twitter.com/orreither) [LinkedIn](https://www.linkedin.com/in/lisaorr) - Engineering Leader, [[Zapier]]
-	- ## Talk: Your Support Team Should Ship Code
+	- ## Talk: Your Support Team Should Ship Code [00:57:46](https://www.youtube.com/watch?v=cMSprbJ95jg&t=3466s) - Zapier (Lisa Orr): Case Study on "Scout," the Support Team AI Agent
 		- [[Zapier]] maintains 8000+ integrations that break as APIs change. We had thousands of backlog support tickets with dozens more arriving weekly. To keep up with the traffic, we started building AI tools to help ship integration fixes faster. We began by shadowing engineers fixing tickets and building tools we believed would expedite the fix process. Our first effort, an API playground hosting AI tools like diagnosis and test generation, failed to get engineering traffic because it pulled builders out of their workflows. We pivoted to MCP tools that engineers could use directly in their IDEs. MCP tools gained traction, but our most valuable tool, Diagnosis, took too long to run. Engineers wouldn't wait for it, revealing we needed an asynchronous approach. We built Scout Agent to string our tools together, autonomously reading support tickets, gathering context, generating fixes with tests, and submitting merge requests ready for review. This agent approach has gained traction with our support team handling high ticket volumes. An MR ready for review means they can validate and ship a fix quickly before needing to jump on the next incoming ticket. Throughout this process we've learned that the real challenge is everything surrounding code generation. Before writing code, Scout Agent needs both the right context and to show its work so engineers trust its recommendations. After generation, engineers need to quickly validate and correct the proposed fix, otherwise MRs sit unreviewed and abandoned. Embedding Scout Agent directly in GitLab solved this. Teams can iterate on proposed solutions without context switching. To track improvement, we measure three distinct failure modes: categorization accuracy (should Scout attempt this ticket?), fixability assessment (does this need a code fix?), and solution quality (does the generated code actually work?). Each reveals different improvement opportunities. Today, Scout drives 40% of support's integration fixes, with expansion to engineering teams and downstream automation (testing, shipping, migration) as our next frontiers.
 	- ## Intro - Grand Canyon
 		- Parallels between grand canyon and zapier: erosion

pages/AI___ES___25___11 Code___LEAD___20 Thu___1005 2026 year ide died steve yegge gene kim amp sourcegraph.md

@@ -1,8 +1,8 @@
-- # 10:05am - 10:24am | AI Leadership | Room: Times Center
+# 10:05am - 10:24am Talk: 2026: The Year the IDE Died
 	- ![Steve Yegge](https://www.ai.engineer/speakers/steve-yegge.jpg) ![Gene Kim](https://www.ai.engineer/speakers/gene-kim.png)
 	- **[[Person/Steve Yegge]]** [Twitter](https://twitter.com/Steve_Yegge) [LinkedIn](https://www.linkedin.com/in/steveyegge) - Engineering Leader, [[Sourcegraph]]/[[Amp]]
 	- **[[Person/Gene Kim]]** [Twitter](https://twitter.com/RealGeneKim) [LinkedIn](https://www.linkedin.com/in/realgenekim) [Website](https://www.realgenekim.me/) - Author & Researcher, IT Revolution
-	- ## Talk: 2026: The Year the IDE Died
+	- ## Talk: 2026: The Year the IDE Died [01:10:31](https://www.youtube.com/watch?v=cMSprbJ95jg&t=4231s) - Steve Yegge & Gene Kim Begin Their High-Energy "Vibe Coding" Talk
 		- As AI has grown more capable, software developers around the world have lagged behind the technology advances, and have consistently eschewed the most powerful tools. In this talk I explore why devs are staying 9-12 months behind the AI curve. I'll share a preview of what 2026's AI coding tools will be like, and paint a vision of where we go from here.
 	- ## [[Person/Steve Yegge]]
 		- ### AI Developer Trajectory — timeline layout

pages/AI___ES___25___11 Code___LEAD___20 Thu___1100 OpenAI Applied AI Bill Chen Brian Fioca Future-Proof Coding Agents.md

@@ -1,8 +1,8 @@
-- # 11:00am - 11:19am | AI Leadership | Room: Times Center
+# 11:00am - 11:19am Talk: Future-Proof Coding Agents: Building Reliable Systems That Outlast Model Cycles
 	- ![Bill Chen](https://www.ai.engineer/speakers/bill-chen.jpg)![Brian Fioca](https://www.ai.engineer/speakers/brian-fioca.jpg)
 	- **[[Person/Bill Chen]]** [LinkedIn](https://www.linkedin.com/in/billchen99/) - Applied AI, [[OpenAI]]
 	- **[[Person/Brian Fioca]]** [Twitter](https://twitter.com/bfioca) [LinkedIn](https://www.linkedin.com/in/brianfioca) [GitHub](https://github.com/bfioca) [Website](https://fioca.com/) - Applied AI, [[OpenAI]]
-	- ## Talk: Future-Proof Coding Agents: Building Reliable Systems That Outlast Model Cycles
+	- ## Talk: Future-Proof Coding Agents: Building Reliable Systems That Outlast Model Cycles [02:11:38](https://www.youtube.com/watch?v=cMSprbJ95jg&t=7898s) - OpenAI Team (Bill Chen & Brian Fioca) Presentation Starts
 		- Coding agents are becoming one of the most active areas in applied AI, yet many teams keep rebuilding fragile infrastructure every time models or providers change. We believe there is a better way. By anchoring on a stable abstraction layer like Codex, we can stop worrying about harness rewrites and focus on the parts of the stack that create lasting value. We treat models as interchangeable sub-agents, plug into shared primitives, and let upstream improvements flow through without breaking products. This lets teams move faster, stay resilient as the ecosystem evolves, and focus their energy on domain-specific workflows and user experience.
 	- ## Agenda
 		- Coding agents

pages/AI___ES___25___11 Code___LEAD___20 Thu___1120 McKinsey Moving away from Agile.md

@@ -1,10 +1,10 @@
 chatgpt-link:: https://chatgpt.com/g/g-p-691f249c2fac8191ab8b4b926da5cb3b-ai-es-25-11-code/c/691f4002-5ff8-800a-875f-b7a02e678f4f
 
-- # 11:20am - 11:39am | AI Leadership | Room: Times Center
+# 11:20am - 11:39am Talk: Moving away from Agile: What's Next?
 	- ![Martin Harrysson](https://www.ai.engineer/speakers/martin-harrysson.png)![Natasha Maniar](https://www.ai.engineer/speakers/natasha-maniar.jpg)
 	- **[[Person/Martin Harrysson]]** [Twitter](https://twitter.com/martinharrysson) [LinkedIn](https://www.linkedin.com/in/martinharrysson) - Partner, [[McKinsey]]
 	- **[[Person/Natasha Maniar]]** - Partner, [[McKinsey]]
-	- ## Talk: Moving away from Agile: What's Next?
+	- ## Talk: Moving away from Agile: What's Next? [02:35:51](https://www.youtube.com/watch?v=cMSprbJ95jg&t=9351s) - McKinsey (Martin Harrysson & Natasha Maniar): Argument that the Agile Operating Model is a Bottleneck
 		- Most enterprises are not capturing much value from AI in software dev to date (at least relative to the potential). The reason is that most are adding AI tools to their dev teams without changing the people and operating model aspects (i.e., limited changes to ways of working, team configurations, role definitions, stage gates, etc.). Many core aspects of software development haven't changed in the past 10+ years, and that's holding us back from moving to the new paradigm of software development! We will share examples of what makes the difference.
 	- ## New technologies have given rise to new software dev methodologies
 	- **Overall visual structure:** A horizontal timeline across the top, spanning **Pre-2000s → 2000s → 2010s → 2020s**, with dots marking each era. Two horizontal rows: **Tech breakthrough** (top row) and **Software dev methodologies** (bottom row). Beneath each era is a large illustrative image capturing the “vibe” of that period’s development culture.

pages/AI___ES___25___11 Code___LEAD___20 Thu___1140 Yegor Denisov-Blanch Stanford AI ROI Software Engineering.md

@@ -1,9 +1,9 @@
 chatgpt-link:: https://chatgpt.com/g/g-p-691f249c2fac8191ab8b4b926da5cb3b-ai-es-25-11-code/c/691f455b-fbe4-800a-be24-adb4242e34ac
 
-- # 11:40am - 11:59am | AI Leadership | Room: Times Center
+# 11:40am - 11:59am Talk: How to Quantify AI ROI in Software Engineering (Stanford Study / 120k Devs)
 	- ![Yegor Denisov-Blanch](https://www.ai.engineer/speakers/yegor-denisov-blanch.jpg)
 	- **[[Person/Yegor Denisov-Blanch]]** [Twitter](https://twitter.com/yegordb) [LinkedIn](https://www.linkedin.com/in/ydenisov) - Researcher, Stanford
-	- ## Talk: How to Quantify AI ROI in Software Engineering (Stanford Study / 120k Devs)
+	- ## Talk: How to Quantify AI ROI in Software Engineering (Stanford Study / 120k Devs) [02:55:05](https://www.youtube.com/watch?v=cMSprbJ95jg&t=10505s) - Stanford (Yegor Denisov-Blanch): Research Shows Codebase Cleanliness Correlates Strongly with AI Productivity
 		- You're investing millions in AI for software engineering. Can you prove it's paying off?
 		- Benchmarks show models can write code, but in enterprise deployments ROI is hard to measure, easy to bias, and often distorted by activity metrics (PR counts, DORA) that say "more" without proving "better."
 		- Drawing on field data from 120k+ developers across 600+ companies, I'll show exactly where AI helps the most and how to measure the ROI of your software engineering AI deployment.