Register: https://www.eventbrite.com.au/e/build-with-ai-30-2026-gdg-brisbane-tickets-1987565006438?aff=oddtdtcreator&keep_tld=true

Join us for an evening of deep dives into the practical side of Generative AI and Data Engineering. We’re moving beyond simple prompts to explore how industry leaders ensure their AI outputs are actually reliable and how serverless agents can automate the "boring" parts of data mapping.

Whether you're a developer working on LLMs or a data engineer looking to leverage Gemini for schema management, this session offers a mix of high-level strategy and hands-on workshop learning.

Agenda

• 5:30 PM – 6:00 PM: Check‑in, Food & Networking (Grab a bite and meet the community!)

• 6:00 PM – 6:45 PM: Talk by Wesley Volz - LLM Evaluation & Lessons from Atlassian RovoDev

• 7:00 PM – 8:30 PM: Workshop by Anupam Phogat - The Self-Curation Engine (Gemini + Cloud Run)

• 8:40 PM – 9:00 PM: Networking & Closing Conversations

Talk by Wesley Volz : It’s easy to make a prompt tweak, but it’s hard to know if you’ve actually improved your model or just shifted the errors elsewhere. Wesley will share a more robust approach to measuring baseline performance. He’ll also dive into specific "lessons learned" from developing RovoDev at Atlassian, showing you how to determine if your tweaks are making outputs more reliable and where you should focus your efforts during development.

Workshop by Anupam: Retail data arrives from disparate sources with inconsistent schemas. To solve this, we present a serverless Cloud Run agent that uses the Gemini API to automate schema mapping. Triggered by new data, the agent compares raw schemas to a BigQuery target and dynamically generates the required SQL or Dataform transformations. This talk highlights how AI automates the most manual parts of data engineering: schema maintenance and logic generation.