Have you ever wanted to release your own product but felt blocked by lack of time, money, or confidence? In this talk I’ll show how Elixir can be used not just to build elegant systems, but to move from idea to traction quickly. I’ll share how I’m building a LiveView-powered learning platform that improves developer efficiency using gamification, cognitive science, and AI - and how a clean Elixir architecture lets me move fast without losing control. This working prototype is built with one clear goal: testing traction. I’ll walk through the technical choices that support rapid iteration, including Phoenix LiveView patterns for interactive flows, structuring multiple Phoenix apps in a monorepo with independent releases, deploying on Fly.io, integrating AI tools like Tidewave, and what I learned upgrading Elixir, Erlang, and OTP while moving quickly. With over 10 years of Elixir experience, I’ll show how good architecture becomes an accelerator, not a blocker, when you’re trying to ship and learn at the same time.
Key Takeaways:
- Attendees will learn:
- How to structure a Phoenix/LiveView project for fast iteration without chaos
- How to organise monorepos and releases for experimental products
- Practical lessons from deploying and upgrading Elixir systems under time pressure
- How AI tools can be integrated into a BEAM application
- A repeatable approach for testing product ideas with Elixir before investing heavily
Target Audience: This talk is aimed at Elixir and Phoenix developers who are building real applications - whether for startups, internal tools, or their own products - and want to ship faster without losing architectural integrity. It’s especially relevant for:
- Developers using Phoenix LiveView to build interactive systems
- Engineers interested in monorepos, releases, and deployment workflows
- Developers exploring AI integration in Elixir applications
- Senior engineers and indie builders who want to test product ideas quickly while keeping their systems maintainable
- Some familiarity with Elixir and Phoenix is assumed, but the talk is accessible to anyone who has built or deployed a BEAM application.