Sparking the Future of AI Security: From AI Observability to Automated Response

Highlights from the AI Tinkerers Hackathon and TLV Meetup

A Weekend of Builders, Agents, and AI Chaos

Three weeks ago, Tel Aviv lit up with the energy of some of the boldest minds in AI for the inaugural AI Tinkerers Hackathon hosted at Microsoft Reactor, which Zenity was proud to sponsor, and the follow-up AI Meetup #1, hosted at AI21 Labs. What started as an experimental gathering of builders turned into a high-energy event full of real demonstrations, late-night commits, and new ideas around AI-native development.

Why Zenity Invests in AI Observability, Detection & Response

As AI rapidly moves from prototype to production, security needs to play a meaningful part in the conversation from day one. Zenity’s mission is to protect organizations embracing AI; whether they’re shipping copilots, internal GPT apps, or multi-agent workflows. These AI community events, both the hackathon and meetup, were attended by hackers and engineers that represent the vanguard of AI innovation, and we wanted to help this community build, customize, and tinker with AI responsibly.

Real-Time AI Monitoring and Security Wins from the Front Lines

• 80+ hackers showed up for a weekend of rapid prototyping, building agents, interacting with developer tools, and incorporating critical AI agent observability and real-time threat detection to strengthen security layers around LLMs.
• Projects ranged from autonomous infrastructure agents to monitoring tools, with several projects directly tackling prompt injection defense, devops tooling, and agent observability.
• The hackathon concluded with live demos and lightning talks, followed by the first AI Tinkerers meetup - a packed room with real builders sharing what they’re working on.

One of the standout moments was Tamir Sharbat’s talk, where he walked through just how easy it is to hack LLM copilots and why that matters in a world where these tools are rapidly becoming part of critical workflows. It wasn’t fear-mongering, but rather a wake-up call.

A Personal Note as the Organizer: Building Community and AI Security Posture Management Together

This event hit close to home.

At Zenity, I spend my days building systems that provide secure AI systems for the world’s leading enterprises, enabling them to build, customize, and use autonomous agents safely—without slowing down business productivity. Outside of work, I am proud to have helped launch the AI Tinkerers community in Tel Aviv; a space for builders who are actually shipping. So when we put together the hackathon and meetup, the goal was clear: bring people together to build ambitious, useful things and think critically about how to protect them.

And that’s exactly what happened.

We need to treat agentic systems with the same rigor we apply to production software. Because they are production software now.

Huge thank you to Zenity for supporting this vision - not just as a sponsor, but as a company that truly understands what it means to secure AI from the inside out and who places customers at the center of everything that we do.

This was just the beginning. Let’s keep building - and not just prototypes. Production-grade, secure-by-design systems.

Leading the Charge on AI Security

Zenity is leading the way in AI security - building real-time defenses, privacy protections, and policy enforcement that move as fast as the models do. In a world of autonomous agents and open-ended prompts, you don’t need more red tape - you need the right guardrails in the right places, before things go off the rails.

We’re not just talking about threats. We’re building AI risk mitigation infrastructure to proactively detect, respond to, and prevent evolving threats—and helping others do the same.

What’s Next

We’re just getting started. The energy in TLV is real - and we’re here to support the builders pushing the boundaries of what’s possible and what’s safe.