Wellington AI Meetup β March 2026
βWellington AI Meetup β March 2026
βTuesday 24th March | 5:00 pm β 7:30 pm
βπ Hnry β 2/45 Johnston Street, Wellington Central, Wellington 6011
βWelcome
βMarch is here and the AI engineering space is moving faster than ever. Personal AI assistants are becoming the new normal, agent swarms are running enterprise security, and fully autonomous code review is already in production β not as demos, but as real systems handling real work.
βThis month Seb from ZunoSmart Labs kicks things off with a look at the rise of personal AI assistants and what we should be paying attention to. Then two talks from Trade Me that you won't want to miss β Kate Pearce on building a 75+-agent security swarm and Joris Decombe on fully autonomous AI code review on every single merge request. The architecture, the failures, and the hard decisions that made it all work.
βGrab a drink, settle in, and come learn from people doing the work. See you there.
βTalks
βSeb Krueger β "OpenClaw and the Rise of Personal AI Assistants"
βPersonal AI assistants are rapidly becoming the new normal β handling our emails, managing our calendars, writing our code, and making decisions on our behalf. The possibilities are genuinely exciting, and we're only scratching the surface. But with great power comes great responsibility. Seb kicks off the evening with a quick look at how assistants like OpenClaw are changing the game, and what we should be keeping an eye on β from data privacy to trust and autonomy β so we can embrace this wave with our eyes wide open.
βπ LinkedIn
βKate Pearce β "Building a 75+-Agent Security Swarm"
βA 5-person security team, 1.5 of whom write agent code in their spare time from their day jobs β running enterprise security at scale through a 60+-agent AI swarm across five levels of hierarchy. Kate shows how Trade Me actually pulled this off, and what fell out of it: a company-wide code review capability that's made security the most popular team among engineers. The architecture, the failure stories, the 81% cost reduction β the internals justify the ambition.
βπ LinkedIn
βJoris Decombe β "Autonomous Code Review at Scale"
βEvery single merge request at Trade Me, reviewed autonomously by AI β no human in the loop. This is the system that made security the team engineers actually want to hear from. Joris built the pipeline: sandboxed Claude Code instances spun up per MR, findings forced into schema, inline comments posted automatically. He then shows everything required to make it safe enough to actually trust: prompt injection defence, sandbox lockdown, timeout chains. The hardest decision wasn't what to give the agent β it was what to withhold.
βπ LinkedIn
βSponsors
βπ’ Venue β Hnry β Best venue in town!
βπ Drinks/Food β Younity β Shoutout to π Joel Norris for being a legend and supporting the AI Meetup!
βπ€ Speakers β Trade Me β Shoutout to the speakers, obviously, but especially to Tracy Morris from Trade Me for helping pull the speakers together for today!