Personalized Agents Hackathon | Hosted by Lightning.AI, Validia

​What happens when Lightning AI's frontier training infrastructure meets Validia's safety & security expertise?

​A hackathon where you build smarter, ship faster, and deploy agents in a secure manner.

​We're excited to host a hackathon right here in New York City, where we'll challenge you and your team to build autonomous agents that work the way you work - faster than ever before.

​What You'll Build With:

​You'll have full access to two powerful resources working in tandem:

​Lightning AI Studios provides the infrastructure to build, train, and deploy your AI agents with the same tools used for frontier model development. Access to serious compute, seamless workflows, and OpenClaw - the framework for building personalized autonomous agents that understand your unique context and workflows.

​Validia's team brings deep expertise in the safety and security behind AI models. We specialize in building detection for unsafe prompts, jailbreak attempts, and distillation attacks against your models.

​The Challenge:

​Build a personalized autonomous agent using Lightning AI's OpenClaw studio that adapts to how you and your team actually work — and does it safely. Your agent should solve a real problem and demonstrate what's possible when you combine cutting-edge AI infrastructure with genuine security built in from the ground up, not bolted on after the fact.

​Think: agents that learn from your unique work patterns, automate repetitive workflows specific to your domain, and orchestrate complex tasks — all while staying resilient against the threats that make real-world AI deployment risky. Can your agent handle adversarial inputs without breaking? Does it know when it's being manipulated or probed? Can it maintain context about how you operate without leaking sensitive information or being distilled by a bad actor?

​The best submissions won't just be the most capable agents. They'll be the ones that are hardest to break.