Beyond Coding

Meta Staff Engineer Sajal Rustagi reveals what Staff Engineers do that most don't and what the best engineers do differently to reach the top. While foundational behaviors like proactivity and ownership still separate growing engineers from stuck ones, leveling up today requires a completely new set of staff skills built around agents.In this video, we cover:The behavioral shift from IC4 and IC5 (Senior) to IC6 (Staff Engineer) at MetaHow top engineers treat on-call firefighting as a system-wide ownership opportunityWhy rotating between Engineering Manager (EM) and Staff IC sharpens your business and technical influenceTransitioning from Developer Experience (DX) to Agent Experience (AX) using MCP, Claude, and markdown contextThis conversation is for software engineers, senior ICs, and tech leads who want to learn the new staff skills required to multiply their impact and get promoted now.This episode was recorded at TechLead Conference, a conference for engineering leaders on adopting AI.Timestamps:00:00:00 - The New Staff Engineer Skills00:00:44 - 2 Behaviors That Separate Mediocre and Staff Engineers00:02:16 - How the Best Engineers Build an Ownership Mindset00:04:26 - What Staff Engineers Do Differently During On-Call00:06:26 - Senior vs. Staff Engineer: Expanding Your Sphere of Influence00:09:56 - Why Switching Between Manager and Staff IC Builds Better Engineers00:14:04 - Why Unblocking Engineers Now Means Unblocking Agents00:16:41 - Managing Multi-Agent Handoffs Without Losing Control00:19:02 - Shifting from Developer Experience to Agent Experience00:21:47 - Why the Empty Engineering Backlog Is Still a Myth00:25:03 - How to Build Custom Agent Workflows and Reusable Skills00:28:31 - The #1 New Skill Software Engineers Must Build Today00:31:10 - How Meta Engineering Culture Compares to StartupsGuest - Sajal Rustagi, Staff Engineer at Meta: https://www.linkedin.com/in/sajalrustagi#StaffEngineer #SoftwareEngineering #EngineeringLeadership

How do senior engineers use coding agents on prod infrastructure when “make no mistakes” is just a polite request? Kerim Satirli (HashiCorp, an IBM company) explains why prompt rules don’t hold in production, and what he runs instead.In this video, we cover:How Infrastructure as Code (Terraform, Contentful, Datadog, Okta, Cloudflare providers) evolved from open-source PRs to automated agent harnessesWhy AI skills files and “make no mistakes” prompts are only requests, and what replaces them in production pipelinesHow AI-generated pull requests are overwhelming open-source maintainers (and why projects like cURL are pushing back)Personalized “software factories”, spec-driven development (GitHub Spec Kit), and cutting AI documentation noise with BLUF (Bottom Line Up Front)What open-source Rust clones of Adobe mean for the future of SaaSFor infrastructure engineers, DevOps practitioners, and software architects who want to move past one-shot prompting and run coding agents on production infrastructure without losing control.Timestamps:00:00:00 - Why “Make No Mistakes” Is a Polite Request00:00:38 - How Open Source Terraform Changed Provisioning00:01:59 - Why API-First Services Matter Beyond AWS and Azure00:04:19 - How AI Agents Changed Terraform Provider Updates00:07:03 - Why “Make No Mistakes” Won’t Stop Your Agent00:10:31 - Bringing Determinism Back to Agentic Workflows00:13:07 - How AI Pull Requests Are Overwhelming Open Source00:15:41 - Why Every Engineer Is Building a Software Factory00:18:52 - Why AI-Generated Documentation Slows Down Humans00:23:12 - Using Agents to Filter Noise and Summarize Specs00:27:02 - Escape Hatches for Automated Terraform Applies00:29:03 - Where the Joy in Engineering Lives Now00:31:24 - Spec-Driven Development and the Return of Architecture00:34:51 - Open-Source Rust Clones, Reverse Engineering, and SaaS00:38:19 - How to Build Software Products People Actually WantGuest: Kerim Satirli, Senior Developer Advocate at HashiCorp, an IBM company.https://www.linkedin.com/in/ksatirli#Terraform #CodingAgents #DevOps

Lead Architect Carlos Kelkboom shares his rules for coding agents and how great engineers architect software with AI. One of the great ones is: he throws away generated code the moment he doesn't understand it. He breaks down why coding agents fail at iterative changes, why his main agent never touches files, and how he scaffolds projects for one-shot execution.In this video, we cover:Why AI coding agents struggle with iterative changes (SlopCodeBench) and excel at one-shot executionHow to manually scaffold code, types, and folder boundaries before letting sub-agents runDesigning a distributed Markdown and indexed-search graph harness for progressive context disclosureUsing Git worktrees to run parallel coding agents and merge the best implementationsWhy dynamic memory and the heap need to disappear in favor of deterministic programming languagesHow AI bubbles and solo coding impact junior engineers and team collaborationThis episode is for software engineers, tech leads, and system architects who want to move beyond "vibe coding" and build maintainable, deterministic production software with AI agents.Connect with Carlos: https://www.linkedin.com/in/carlos-k-4016b8225Timestamps:00:00:00 - Why Carlos Throws Away Working AI Code00:00:53 - Why Pull Requests and Team Collaboration Are Dying00:03:50 - The Trap of Starting Projects You Never Finish00:06:06 - Escaping the Solo AI Developer Bubble00:11:54 - How to Scaffold Code Before Using AI Agents00:15:54 - Learning New Programming Languages With AI Agents00:21:14 - Why We Must Get Rid of the Heap00:28:32 - Building a Shared Language Across Engineering Teams00:31:31 - Vibe Coding vs Deep Software Engineering00:40:12 - Code Is Cheap: Git Worktrees for Parallel Agents00:41:45 - How Deep to Scaffold Before the Agent Runs00:45:09 - Democratizing Software, Missing Juniors, and the AI Bubble00:53:33 - Designing a Graph Harness for AI Agent Context00:58:18 - Why His Main Agent Never Touches Files01:03:07 - Throw Away Code You Don't Understand01:06:25 - Rethinking Code Reviews and Mutation Testing for AI#CodingAgents #SoftwareEngineering #AgenticEngineering

What happens when coding agents see your entire system: code, services, infrastructure and live traffic, all connected? Most agents are blind to everything outside the repo. Adyen Staff Engineer Mat Jones explains what changed once theirs weren't, and why he calls it somewhat scary.In this video, we cover:Running five or six coding agents in parallel with git worktrees and Atrium, an internal tool that lets agents coordinate and share memoryArchitect: nine connected graphs in Neo4j covering services, code, DNS, data stores and infrastructureIngesting 7 million trace spans a second with Kafka and Apache Flink, and what collapsed on the first trySentinel: anti-pattern monitoring, with agents that raise the fix as a merge requestMCP vs APIs vs CLI tools, and how to deal with tool bloatThis episode is for software engineers and platform teams who want coding agents to reason about real systems, not just the repo in front of them. You'll leave with a concrete place to start and an honest view of the risks.00:00:00 - Intro00:00:42 - Running Five Coding Agents in Parallel00:01:35 - Atrium: How Agents Avoid Stepping on Each Other00:06:57 - Why Graphs Save Coding Agents Tokens00:12:25 - What Is Architect? Nine Graphs in One Database00:16:54 - What Broke at 7 Million Spans a Second00:20:37 - The Timeline From Proof of Concept to Production00:27:59 - Tracing Code Changes to Critical Payment Flows00:30:26 - How to Start: Build Scrappy, Then Earn Buy-In00:35:07 - Making a Codebase Agent Friendly With Skills and Hooks00:42:07 - Why You Should Start With a Service Graph00:44:46 - The Downside of Everyone Seeing Everything00:48:34 - Sentinel: Agents That Find and Fix Anti-Patterns00:53:01 - Are We Going Too Fast With AI?01:01:55 - Two Engineers, Four Months: Why Big Bets Got Cheaper01:07:11 - MCP vs CLI Tools and the Tool Bloat ProblemGuest: Mat Jones, Staff Engineer at Adyenhttps://www.linkedin.com/in/mathewwjones#codingagents #softwareengineering #platformengineering

HumanLayer CEO Dexter Horthy, who coined "context engineering", on why AI agents degrade codebases over time. He breaks down a benchmark that measures AI code quality across features, and how his team ships with agents without drowning in slop.In this video, we cover:SlopCodeBench, SWE-bench and which coding benchmarks to trustToken maxing, dark factories and the "Instagram phase" of AI engineeringWhy 2-4x productivity beats chasing 100xPull requests, design docs and how engineers build taste nowDex's agentic engineering workflow: one-shots, overnight janitor bots and research-design-planFor engineers and engineering leaders adopting AI coding agents who want to know what's real and what works in teams at scale.Timestamps:00:00:00 - The Man Who Coined the Term: Context Engineering00:00:54 - SlopCodeBench: Proof AI Makes Codebases Worse Over Time00:05:49 - Background Agents and the Enterprise Friction Problem00:08:13 - Why You'll Never One-Shot a Big Feature00:11:18 - Scaling Agentic Engineering to 1,000 Engineers00:16:19 - Dark Factories, Fake Hype and Token Maxing00:19:55 - Why 2-4x Faster Beats Chasing 100x00:21:31 - Should Teams Stop Doing Pull Requests?00:24:20 - Can Deterministic Tools Replace Code Review?00:27:37 - How to Build Engineering Taste When AI Writes Code00:33:05 - Dex's Agentic Workflow: One-Shots and Janitor Bots00:35:18 - Research, Design, Plan: Shipping Big Features With Agents00:38:35 - Why Overnight Agents and Heavy Parallelism Backfire00:40:10 - Subscription Maxing vs Paying Per Token00:42:40 - Parallelize Like Poker, Not a Slot MachineDex Horthy: https://x.com/dexhorthyDex's talks and deep dives: https://aithatworks.bold.video#AgenticEngineering #AICoding #SoftwareEngineering