Revisit Old Problems with New Eyes!
summary
Argues agents let one engineer build and ship things that used to need deep pockets: Taxly deployed in under two weeks is a VAT‑filing app with ~85% test coverage and full check suites; an MCP server to surface PR comments in the developer’s local VS Code chat terminal; and a GitHub Action that uses the Anthropic SDK for agentic semantic versioning that analyzes PR diffs to label breaking/minor/patch before merge. Shows trunk‑based + feature‑flag deployments, canary/beta targets, and instant soft rollbacks.
AI-written summary · about
chapters (31)
- 0:00Revisit Old Problems With New Eyes
- 1:10Speaker Intro - Pramod Yadav, Test Engineering Manager, ABN AMRO
- 2:13Building a VAT Filing Product, a PR Review MCP Server and More
- 2:45The Constraints That Existed Last Year Simply Don't Exist Now
- 3:17Live Poll: How Well Do You Understand Continuous Delivery?
- 5:26The Continuous Delivery Diagram: Six Swim Lanes Explained
- 7:29The Planning Phase and Where Feature Flags Fit In
- 9:34The Commit Phase: Pre-Commit Hooks, Linting, Affected Tests
- 11:38Semantic Versioning and Why Existing Solutions Are Fragile
- 13:11Agentic Semantic Versioning: Letting an Agent Judge Breaking Changes
- 14:44Why That Call Belongs Inside the Pull Request, Not After the Merge
- 15:14Continuous Integration: Checks, Coverage Gates and Merge Policies
- 16:15Taxly: A Real Product With 85% Coverage and Full Check Suites
- 16:45Skills, MCP Servers and Test Layers Inside a Real Repo
- 18:17Continuous Delivery: Building the Image and Deploying to Test
- 19:19Using Agents for Exploratory Testing - and Being Impressed
- 20:21Deploying to Production With the Feature Flag Off
- 21:22Dark Launching and Shrinking the Blast Radius
- 21:52Canary Releases, Beta Groups and A/B Experiments
- 22:23Instant Soft Rollbacks Instead of Hot Fixes
- 23:24When It Isn't Behind a Flag: Rollback vs Hotfix Branch
- 24:26TaskFlow: Demonstrating Trunk-Based Development With Feature Toggles
- 25:27The Same Commit Behaving Differently in CI, Staging and Production
- 27:02Repos You Can Use to Demonstrate These Concepts Yourself
- 28:35Roles May Break Vertically Instead of Horizontally
- 31:09Q&A: How AI Helps Without Overreaching
- 33:13Q&A: Feature Flags vs Short-Lived Branches
- 34:48Q&A: The Real Risk of Over-Reliance and Skill Atrophy
- 36:22Q&A: Which Old Testing Problem AI Is Finally Cracking
- 37:55Q&A: Test Data, Environments and Flakiness - Which to Attack First
- 39:28Q&A: Who Is Accountable When an Autonomous Agent Causes an Incident
related talks
24mIt's Time To Rethink Everything
17mSoftware Development Agents: What Works and What Doesn't
18mHow I shipped more code and products than ever before
06h 22mUsing AI to Enhance Developer Productivity
30mBuilding Tools with Tools: Modern AI-Driven Development
55mKeynote: Exoskeletons, not Autopilots
46mThe amplified developer: zero to shipped in 30 mins with AI Coding Ass
12mHow to maintain a codebase – when everyone can code
