AI in Practice: Agent → Secure → Industry
Description
Over three Sunday afternoons at Swinburne University you'll go from the foundations of context and skills — the pieces that actually make an AI agent work — to knowing what's safe to put into these systems and how to apply AI effectively inside a real business. Every session is hands-on: bring a laptop, leave with an agent you can use on Monday.
Come to one, or come to all three. They build on each other.
WHAT'S INCLUDED
· Attendee certificate from AVA Training
· 7 days of post-workshop support on Slack — ask questions while you're actually applying it
· Support in English, Chinese, and German
WHAT TO BRING
· Your own laptop
· A Claude or ChatGPT subscription, so you can use Claude Code / Codex on the day
WHERE
Melbourne CBD. The exact venue is announced closer to the date — ticket holders are notified by email.
PRESENTED BY
AVA Training, with Unwind AI and Great Monkey Lab.
Agenda
- 1:00 pm - 4:00 pm
Workshop 1 · Start Here: Build Your First AI Agent — Sunday 16 August
Walk in curious, walk out with a working agent. What's happening inside a model when it answers you — and why that changes how you structure your promt and supporting cotext and skills. Hands-on with Claude Code or Codex side by side, so you can feel where they differ. You'll design an agent profile around your own work, give it skills, and leave with a starter kit you'll keep building on. Presented by Eva Ng, Great Monkey Lab.
- 1:00 pm - 4:00 pm
Workshop 2 · Build Safe: Protect Your AI from Prompt Injection and Data Leaks — Sunday 23 August
Use AI without handing over your data. Why models make things up, and what that means for anything you trust them with. Break a chatbot with prompt injection — legally, in the room — then learn to spot the same attack hidden inside documents and emails. Five layers of practical defence, and you'll leave with a 10-point checklist you can use immediately. Presented by Sowat Rafi, Unwind AI.
- 1:00 pm - 4:00 pm
Workshop 3 · Go Live: How AI Actually Works in Industry — Sunday 30 August
Most AI pilots never make it into the business. This session is about the ones that do. Why the data platform underneath decides whether an agent works or quietly returns the wrong number. Where AI genuinely pays for itself, and where it just looks impressive in a demo. Implementation workflows that survive contact with a real company — and the technical and organisational constraints that make projects succeed, stall, or fail. Presented by Thomas Taresch, Unwind AI.
Lineup
Eva Ng
Eva is an AI Engineer, educator, and founder of Great Monkey Lab. She holds a Master of Data Science from the University of Melbourne and has built applied AI systems across healthcare, energy, engineering, and business automation. Her experience spans AI agents, multimodal systems, cybersecurity, and cloud deployment. Through her workshops, she helps participants turn AI concepts into practical solutions they can confidently build and apply.
Sowat Rafi
Rafi is an AI Engineer and Cybersecurity professional specialising in computer vision, video analytics, and secure AI systems. He has contributed to the development of AI-powered video analysis solutions for physical security and holds a Master’s degree in Cybersecurity from La Trobe University. He is also the creator of Sanket, an offline communication platform designed for network outages and emergency situations. His workshops focus on practical, real-world applications of AI, cybersecurity, and emerging technologies.
Thomas Taresch
Thomas Taresch builds the data platforms and AI agents that decide whether an AI project delivers or quietly stalls. He is the founder of Unwind AI and Data, where he designs data pipelines, AI agents and automation for manufacturers and small businesses across Australia. His path started with an apprenticeship at BMW, followed by engineering studies in Munich and Wales, and twenty years in quality engineering across Germany, Wales and Australia — most of it at Ford Motor Company, where he works as a problem-solving specialist and holds Six Sigma Master Black Belt certification. His machine learning and data automation projects have saved up to $5 million a year and won global innovation awards.
Tickets for good, not greed Humanitix dedicates 100% of profits from booking fees to charity
