Joel Clermont has been building software since discovering a love of computers as a kid and has worked professionally with PHP since the mid-1990s.
Along the way, he has shipped production applications across multiple languages and ecosystems, including Elixir, Erlang, .NET, and Elm. Those experiences have given him a perspective that occasionally challenges conventional Laravel wisdom, grounded in years of hands-on experience across multiple ecosystems.
Joel is known for his thoughtful, approachable style and enjoys engaging in spirited discussions about software development and engineering practices.
Joel Clermont has been building software since discovering a love of computers as a kid and has worked professionally with PHP since the mid-1990s.
Along the way, he has shipped production applications across multiple languages and ecosystems, including Elixir, Erlang, .NET, and Elm. Those experiences have given him a perspective that occasionally challenges conventional Laravel wisdom, grounded in years of hands-on experience across multiple ecosystems.
Joel is known for his thoughtful, approachable style and enjoys engaging in spirited discussions about software development and engineering practices.
Your tests pass. Coverage looks healthy. Yet somehow a production bug still slips through.
Whether those tests were written by you, inherited from a teammate, or generated by AI, the question remains the same: are they actually proving anything, or simply providing reassurance?
In this talk, Joel explores the principles and tooling he uses to answer that question, including an objective way to measure what a test suite is genuinely capable of catching.
The talk begins with a common misconception: code coverage shows what code executed during a test run, not whether a test would actually detect a defect.
From there, you'll explore mutation testing in Laravel, a technique that deliberately introduces defects into an application and measures whether the test suite notices. Through practical examples, Joel demonstrates the kinds of weaknesses mutation testing can expose, including gaps that frequently appear in both hand-written and AI-generated tests.
The talk also examines the trade-offs involved, when mutation testing is worth the investment, and how to focus effort where it provides the greatest value.
You'll leave with a practical workflow for evaluating any test suite: review it using clear criteria, validate it with mutation testing, and strengthen the areas that matter most.
You'll also gain a framework for assessing AI-generated tests with the same rigour you would apply to a junior developer's pull request. Most importantly, you'll begin thinking about tests differently - not "do they pass?" but "if I introduced a bug right now, would they catch it?"
Intermediate Laravel developers who write tests - or have AI writing them - and want greater confidence that their test suites are actually protecting their applications.
Joel Clermont
Owner
@No Compromises
Joel Clermont has been building software since discovering a love of computers as a kid and has worked professionally with PHP since the mid-1990s.
Along the way, he has shipped production applications across multiple languages and ecosystems, including Elixir, Erlang, .NET, and Elm. Those experiences have given him a perspective that occasionally challenges conventional Laravel wisdom, grounded in years of hands-on experience across multiple ecosystems.
Joel is known for his thoughtful, approachable style and enjoys engaging in spirited discussions about software development and engineering practices.
Subscribe for regular updates on the Laracon AU 2026 conference
BECOME A SPONSOR
For 2026, Laracon Australia is offering even more ways for companies and open source projects to sponsor this year's event. Enter your details and we'll be in touch with our 2026 sponsorship prospectus, or contact us at sponsors@laracon.au if you have any questions.