Stop Prompting, Start Delegating: See the World's Most Powerful AI Agent Live
Presented by
Head of Engineering at Findable
The Era of Chatbots Is Over
OpenClaw has taken the world by storm — 224,000 GitHub stars, Mac Mini shortages, and security headlines from Kaspersky to Wired. Everyone’s talking about it. But almost nobody is showing how to use it safely at work.
In this session, Head of Engineering Johannes Berggren shares how Findable went from experimenting on a laptop to running a fleet of AI agents in production — and gives you a practical framework for doing the same.
What You’ll Learn
What is OpenClaw, and why does it matter?
A quick primer on the open-source AI agent that’s redefining how teams work — from chatbot to digital operator.
Real use cases running in production
See what Findable actually uses OpenClaw for: autonomous bug resolution, automated PR reviews, HubSpot triage, and a multi-agent pipeline where bots hand off work to each other — with humans approving only the final step.
The security conversation — honestly
We’ll address the Kaspersky, Wired, and Microsoft headlines head-on. What are the real risks? Why do they exist? And why most of them come from misconfiguration, not the tool itself.
From Playground to Production
Findable’s three-tier deployment model:
- Playground — personal laptop, great for learning
- Dedicated — clean-installed machine for team pilots
- Production — Docker containers with fleet management, isolation, and centralized control
We’ll show the actual architecture: container isolation, reverse proxy, scoped API access — the same patterns your IT team already uses for microservices.
The Trust Ladder
A five-step framework for gradually increasing agent autonomy:
- Read — observe only
- Summarize — digest and report
- Draft — prepare actions for review
- Act + Approval — execute with human sign-off
- Autonomous — full delegation
Most professional setups live at levels 3-4. This is what you take back to your IT department.
Live Demo
Watch a production OpenClaw setup in action — with safety controls enabled and the Trust Ladder applied.
Who Should Attend
- CTOs and engineering leaders evaluating AI agents for their teams
- Developers who’ve tried OpenClaw and want to go from hobby to production
- IT and security professionals who need to understand the risk model
- Anyone curious about using AI agents safely in a professional setting
- People who want to know what to tell their IT department
No deep technical background required — we’ll cover both the strategic and practical angles.
About the Speaker
Johannes “JB” Berggren is Head of Engineering at Findable, where he leads a team building AI-powered building document management for 150+ property owners across Europe. He co-founded Learnlink (edtech), invests in early-stage tech companies through CTO Roundtable Invest, and DJs at Futoria in Oslo. He’s been deploying OpenClaw agents in production since day one.
Connect with JB on LinkedIn.