Testing Gen AI with Paul Maxwell-Walters & Influence in Testing by Rashik Prasad

A bumper double-header of talks for our July Event with Macquarie Group's Rashik Prasad and Sydney Testers' own Paul Maxwell-Walters (and graciously hosted by Talenza), with the keynote covering how testers can achieve influence in decision making by Rashik Prasad and the next talk about testing Gen AI system.

Talk 1: How Your Teams can Test And Evaluate Gen AI and the Software that Uses It - Paul Maxwell-Walters

There has been much talk in QA about using Gen AI tools in our work to generate test cases and automation – however until recently very little on how testers can test the non-deterministic outputs of AI and evaluate their quality. Typically, direct testing of Gen AI is ignored and left for AI engineers and data scientists.

This is a grave mistake. As products using Gen AI like ChatGPT become mainstream and used by all, testers need to learn how to write and automate testing of them – and fast. My talk is on techniques, tools and workflows used to validate the outputs of AI models and agents – including correct prompting, semantic and sentiment analysis, standard outputs, evals and LLM-as-Judge. I will call on my experience as QA lead at a Gen AI tech-ed startup , along with other approaches and use cases, to give info, tips and one or more demos to help you and your teams meet the challenge of testing Gen AI-powered applications.

Talk 2 (Keynote): Technical Excellent Alone Never Guarantees Success, Rashik Prasad

Technical excellence is the foundation of every successful testing career.

*It's what we're taught to develop.

• It's what we spend years refining.
  *And it's what our profession rightly values.

But after leading testing across some of Australia's largest enterprise transformation and regulatory programmes, I've come to realise something that fundamentally changed the way I think about testing. Technical excellence alone never guarantees success.

• Why do technically exceptional testing professionals sometimes struggle to influence important decisions?
  *Why can the same testing evidence lead different stakeholders to completely different conclusions?
• And why do some testing professionals consistently help programmes move forward, even when the testing itself hasn't changed?

In this thought-provoking keynote, Rashik challenges one of the profession's most deeply held assumptions and explores what he believes is the missing half of testing.

Drawing on real enterprise experiences, you'll discover why the role of testing doesn't end when the evidence is produced, and why the most effective testing professionals understand how testing helps organisations make better decisions.

This isn't a session about becoming less technical. It's about understanding how technical excellence achieves the influence it deserves.

Whether you're a Test Analyst, Test Lead, Test Manager or senior quality leader, you'll recognise many of the situations discussed throughout the session because you've probably experienced them yourself.

You may not leave agreeing with every idea. But I suspect you'll never look at testing in quite the same way again.

--------------------------------------------------------------------------

There will be the usual opposrtunities for networking and fun and some prizes to give away! Not to be missed!