Rangle

EventsWebinar

What's a Developer's Job When Agents Do the Building?

Agents can now build and test code faster than people can review it. Rangle Solutions Architect Abdella Ali shares what that means for developers, and the skills that matter more now: spotting when the work is wrong, knowing where people need to step in, and keeping the whole process moving.

  • Tuesday, December 15, 2026
  • 12:30–1:30 PM ET
  • Virtual · Live

Save your seat

Reserve your spot for the live session.

What you'll get

  • The full one-hour session, live, with time for your questions.
  • The complete session recording, sent to every registrant.
  • A calendar hold with your join link.

Can't join live? Register anyway. Everyone who signs up gets the recording.

What's a Developer's Job When Agents Do the Building? Rangle webinar with Abdella Ali
NikeStripeGoogleSanofiStaplesThe North FaceKenvueHims & HersInvescoEndy

When agents do the building, the developer's job moves one layer up.

Two days for a pull request review used to be fast. Next to agents that write a feature in an hour, it's now the slowest step in delivery, so it's the next thing everyone will try to remove.

At the same time, a growing share of your users will show up through agents of their own. Every incentive pushes toward tighter loops with fewer people in them. Someone still has to decide where a person stays, and what they need to see when they do. That's the developer's job now, one layer up.

Abdella Ali is a Solutions Architect with more than a decade at Rangle. This year he has been building agent pipelines for real client delivery work. He first gave this talk at FITC Web Unleashed in Toronto on October 27.

What you'll leave with

  • A clear read on the trend lines. How long agents can work alone today, how fast that's changing, and what it suggests for 2027, clearly labelled as extrapolation.
  • A way to see where your loop waits on people. Checking step by step, task by task, or by outcome, and why review has become the bottleneck.
  • Loops that report to people. Designing architecture reviews, debt audits and security checks that tell the right person exactly what they need to decide.
  • Hard questions for your codebase and Figma file. What replaces the pull request? How do you build a design system agents can't break? What happens to your homepage when the visitor is an agent?

What we'll walk through

  1. Where we've always sat, and how each layer of tooling made each loop do more
  2. Why the loop is tightening, with the data
  3. A real loop: an agent-driven site migration of about 100 pages
  4. Where the loop hits a person
  5. Everyone else gets agents too, and the four actions where they still stop and ask: send, pay, sign, delete
  6. What we bring, one layer up: knowing when it's wrong, deciding where people sit, and keeping the loop tight

Then Q&A.

This session is for you if…

  • You're a developer working out where your own judgement matters most as agents take on more of the building.
  • You're an engineering manager, tech lead or architect whose team already hands real work to agents.
  • You lead design or front-end work and you're wondering what a design system looks like when agents assemble the screens.
  • You feel review becoming the slowest part of delivery and want a better way to decide where people stay in the loop.

Bring your team.