Beyond Coding

How do you prove AI is shipping more features? Amos Haviv leads the Developer Workflow teams at Booking.com, supporting 4000 engineers operating 8000 repos.
Everybody is burning through their AI budget right now and almost nobody can answer what it bought them. Amos can, because his team spent four years building an event system to debug their own SDLC before AI upped the urgency.
In this video, we cover:
Why verification is the bottleneck right now, and where it moves next
Building an event store that separates KTLO from real feature delivery
Why static dashboards create the metric they measure, and the cobra story behind it
Agent cost, model routing, and why Booking ignores token maxing entirely
Running a developer survey with a 92% response rate across 3k+ engineers
Who should own skills and MCPs: a central platform team or the domain experts?
For platform engineers, engineering leaders, and anyone being asked to prove ROI on AI tooling this quarter.

Timestamps:
00:00:00 - Everyone is burning through their budget
00:00:32 - Verification Is the Bottleneck Every Team Hit
00:03:35 - 4,000 Engineers and 8,000 Repos at Booking.com
00:06:48 - Why Copying Google and OpenAI Will Break You
00:09:21 - Verification Is a Stack of Agents, Not One Review
00:13:27 - Cost Is Becoming a Bottleneck of Its Own
00:17:14 - Was the Internet a Bubble? What That Teaches Us
00:25:32 - What Working With the Frontier Labs Looks Like
00:28:26 - Debugging the SDLC With Four Years of Event Data
00:30:24 - Do Engineers Using AI Actually Ship More Features?
00:37:13 - Where to Start If You Measure Nothing Today
00:45:01 - The Cobra Effect: When a Metric Becomes a Target
00:52:23 - Everyone Is a Builder Now, and Everything Needs Support
01:01:21 - Is AI Turning Every Engineer Into a Manager?
01:03:46 - The Developer Survey With a 92% Response Rate
01:10:09 - Who Owns Skills, MCPs, and the Enterprise Harness
01:17:46 - Great Developer Experience Is High Velocity

Mentioned in the episode:
High Output Management by Andy Grove
The Sovereign Individual (1997)
The story of General Magic

Views expressed are Amos's own and do not represent Booking.com.
#AI #SoftwareEngineering #DeveloperExperience

"I need to stop using Opus. This doesn't work." That was Heitor Lessa's conclusion after a refactor cost him 200 million tokens, and it forced him to rebuild the entire agent workflow now available for 1400 engineers. Heitor spent 11 years at AWS, built Lambda Powertools to 230 billion API calls a week, and in this episode he walks through the full SDLC workflow on screen, from discovery to merge check.
In this episode, we cover:
The product loop: discovery, whiteboarding, and the /roadmap command
Spec-driven development with Open Spec and why vanilla setups fail
Three model tiers: SOTA for planning, mid-tier for implementation, cheap models for reviews
Merge checks with adversarial reviewers and attestations that catch agents fabricating test results
The /retro command: using the Socratic method to make your workflow more deterministic
If you're an engineer figuring out how to work with agents at team scale without losing trust in your codebase, this is the workflow to steal. This is also the first Beyond Coding episode with visuals on screen, so let me know what you think of the format.
Timestamps:
00:00:00 - The Math Doesn't Add Up
00:00:43 - Amazon Hypergrowth: 11 Years, 8 Different Roles
00:03:29 - Learning From the Trenches as a Technical Account Manager
00:08:38 - Developer Identity and the Birth of Lambda Powertools
00:10:20 - The Hard Parts of Working in Public
00:13:12 - How Powertools Hit 230 Billion API Calls a Week
00:16:42 - Career Advice: Learn Adjacent Roles, Not More Tech
00:19:37 - When Leadership Decisions Don't Make Sense to You
00:23:21 - The Product Loop Starts With Discovery
00:25:22 - From Whiteboard to /roadmap
00:27:37 - Why Humans Plan First and Agents Come Second
00:30:33 - Commands vs Skills Across 32 Different Models
00:33:38 - Adversarial Reviewers on Every Plan
00:36:07 - The Socratic Method, Explained
00:40:29 - Why He Only Takes Paper Notes
00:44:43 - The Five-Line Paper Trick for High-Stakes Meetings
00:48:18 - /new-work: Capturing Scope Creep Without Derailing
00:54:03 - The Dev Loop Begins: Open Spec Explore
00:56:34 - Three Model Tiers: SOTA, Mid, Cheap
00:57:43 - The $5,000/Month Per Engineer Question
00:58:57 - Guardrails vs Autonomy for 1,400 Engineers
01:04:22 - Auto-Sizer: Does This Task Even Need a Spec?
01:07:26 - Decision Fatigue and Why Frameworks Win
01:09:10 - The Plan Phase: Specs, Design, Formal Verification
01:13:07 - The Refactor That Cost 200 Million Tokens
01:15:11 - When Agents Forge Evidence They Ran Your Tests
01:17:27 - Local-First Architecture Explained
01:23:04 - The Apply Phase: Fully Autonomous Loops
01:24:30 - Coding Was Never the Bottleneck
01:26:39 - Why This Workflow Is an Investment
01:27:39 - Decision Logs and the /onboarding Command
01:29:06 - Running Agents Locally With Enterprise Governance
01:32:42 - Hooks: Making Quality Gates Deterministic
01:36:02 - Merge Checks: 15 Adversarial Reviewers Per Change
01:38:30 - /retro: Interviewing Yourself to Improve the Loop
01:43:12 - Trust, Loss of Trust, and Recovery With Agents
01:48:02 - Experience, Scars, and Critical Thinking
01:49:32 - Why Right Now Is the Time to Experiment
01:52:04 - Conviction Comes From Being in the Loop
#softwareengineering #aiagents #aws

What senior engineers do differently has less to do with output than most career ladders suggest, and Lindsey Simon, VP of Engineering at Vercel, has watched the distinction sharpen as everyone in the valley becomes a "member of technical staff." From why new grads with hackathon years might out-prepare engineers with six years on the job, to what happens when PR throughput stops being your lever, this is a conversation about what earns seniority now.
In this episode, we cover:
Why engineering roles are consolidating into "member of technical staff"
How to ask agents first and frame better questions to humans
The scope-of-impact ladder and what the best engineers systematize
Learning how to learn: closing gaps to 100% understanding
Why writing is the skill that scales
If you're wondering whether your years of experience still compound, or you're early-career and tired of the "woe is the juniors" narrative, this one reframes both.

TIMESTAMPS
00:00:00 - Impact the Business
00:00:31 - FOMO all the time: The 2006 Google Interview
00:01:52 - Engineering Roles Are Consolidating
00:02:52 - The "Member of Technical Staff" trend in SF
00:03:33 - Interns Demo to the CTO
00:04:33 - How New Grads Out-Prepare Senior Engineers
00:06:10 - Ask Your Agent Before You Ask a Human
00:08:01 - Digging Backwards Into Fundamental Understanding
00:09:22 - "We're All Junior Engineers Again"
00:10:22 - Management Is Not Leadership
00:12:04 - Losing PR Throughput as Your #1 Lever
00:13:11 - Fulfillment Beyond Shipping Features
00:14:32 - Building for Fickle Engineers: Telemetry Beats Opinions
00:16:11 - Watching Users Struggle With Your Product
00:18:12 - Have Expectations for Seniors Actually Changed?
00:19:57 - Claude Says a Month, It Takes Two Hours
00:20:33 - What the Best Engineers Do Differently
00:21:30 - How Vercel React Skill Came to Be
00:22:23 - Why Conference Conversations Hit Different
00:23:35 - Learning How to Learn: Close Gaps to 100%
00:25:41 - The Case for Liberal Arts in Tech
00:27:03 - Get Feedback Early, Don't Hide in the Cave

Guest - Lindsey Simon, VP of Engineering at Vercel:
https://www.linkedin.com/in/lindseysimon
#softwareengineering #ai #careergrowth

The fastest engineers are falling behind, and Kitze was one of them. He built his reputation on raw coding speed, then realized his coding wasn't competing with anyone's coding anymore, it was competing with their setups. Wake-up call for developers: Kitze now runs 140 projects solo with agent loops, and in this episode he breaks down what separates the engineers pulling ahead from the ones getting left behind.
In this episode, we cover:
Vibe coding vs vibe engineering, and how to get better results from your agents
Police files: Self-correcting loops that end every agent turn with zero errors
Why teams of 10 are collapsing into teams of 2, and who survives
The rude awakening coming for engineers who refuse to adapt
The number one advice to stay on track and fight FOMO
For individual contributors, tech leads, and principal engineers who don't plan on falling behind

TIMESTAMPS:
00:00:00 - Intro
00:00:40 - Vibe Coding vs Vibe Engineering: The Real Difference
00:02:16 - Police Files: The Self-Correcting Loop on Every Turn
00:05:23 - Capture Every Frustration as a Rule
00:06:53 - Why Being the Fastest Coder Stopped Mattering
00:09:45 - Problem Solver vs Problem Lover: Pick One
00:10:43 - The Rude Awakening Engineers Don't Want
00:12:05 - Why Teams of 10 Become Teams of 2
00:13:09 - Loop Engineering: The Edge Anyone Can Build
00:16:08 - Why No Agent Orchestrator Works Yet
00:17:07 - Starting a Fresh Codebase: What Kitze Transfers
00:19:14 - No Sidebars: Inventing an Agentic OS
00:21:07 - How Kitze Shipped 300 Changes Across 200 Repos
00:23:40 - We Are Becoming the Bottleneck
00:24:25 - Why Leadership Must Give Engineers Room to Experiment
00:26:16 - The Token Divide: Not Everyone Can Compete
00:27:41 - Learn Now or Lose Access Later
00:29:33 - The Culling: Coasting Is Going Away
00:30:45 - Why LLM Code Reviews Beat Tired Seniors
00:33:21 - Solo Engineers With Agent Swarms vs Teams
00:34:53 - Agents Climbing the Org Chart to CEO
00:36:03 - What Distinguishes the Best Engineers: Unblocking
00:36:50 - Ego Is the Real Bottleneck
00:37:55 - Kitze's #1 Advice: Stick to One Model

Danila Shtan runs engineering at Nebius, one of the biggest AI clouds in the world, and he told me exactly which engineers he hires on the spot. There are only hundreds of people on the planet with the skill he wants most, and it is not the one you are grinding on. We get into which engineering skills are actually scarce and well paid today, and which ones are quietly on the way out.
In this episode we cover:
The engineering skills in highest demand right now and which ones are on the way out
Why an AI cloud CTO restricts Claude Code inside his own company
Dan's rule for merging any AI-written code into production
Why working with an agent is like managing a junior engineer
The interview question that surfaces top tier engineer qualities
Why he still runs algorithm interviews today
If you are an engineer trying to work out where the value sits now that agents write the easy code, this is a straight answer from the person building the infrastructure underneath all of it.

Timestamps:
00:00:00 - AI Agents doing everything is a lie
00:00:44 - What Nebius Actually Does
00:04:31 - The Engineers In Highest Demand Right Now
00:06:58 - Inside the Hiring Process
00:08:12 - The Bootcamp: You Join the Company, Not a Team
00:10:51 - Why You Can't Use AI in Their Interviews
00:16:31 - Why He Banned the Word "Headcount"
00:22:25 - Why a CTO Is Not a Technical Role
00:24:49 - The One Skill Every Manager Needs
00:25:48 - Why Smart People Fail at This
00:28:17 - "The Promise of Agents Is Bullshit"
00:31:39 - How AI Multiplies Your Baseline Skill
00:35:32 - Why an AI Agent Is Just a Junior Engineer
00:36:57 - Why He Won't Let His Team Use Claude Code
00:37:46 - His Rule for Merging AI-Written Code
00:40:28 - The Interview That Predicts Great Engineers
00:42:32 - From T-Shaped to Round-Shaped Engineers
00:44:30 - Is There Still a Path for Juniors?
00:45:28 - Why Hard Skills No Longer Matter
00:47:11 - The Engineers Who Will Become Obsolete
00:50:04 - The Real Reason People Stay at Banks
00:52:26 - Where AI Agents Actually Help
00:54:40 - Why He Still Uses Algorithm Interviews
00:56:05 - Tech Enthusiasts vs. Real Engineers
#AIEngineering #TechCareers #SoftwareEngineering