The alert that started everything.
2:30 AM. Phone buzzing. WannaCry ransomware had hit, and one of us was inside Microsoft, patching and securing internal machines. No playbook. No automation. Just a terminal, a growing list of affected hosts, and a lot of coffee.
By 4:30 AM the immediate threat was contained. By 9:30 AM, it was international news.
A simple question to a friend. That conversation never stopped.
The idea that wouldn't go away.
Life moved on. Careers grew. One of us moved to Melbourne. Another stayed at Microsoft in Redmond. But every few months, the same conversation would come back: why are engineers still doing the same manual toil during incidents? Why hasn't someone automated the boring, repetitive parts of being on-call?
The honest answer was that the tech wasn't ready. Language models couldn't reason through multi-hop dependency graphs. Knowledge retrieval wasn't good enough. Agentic architectures were still academic papers, not production systems.
So we waited. And we watched. And we kept talking.
The tech finally caught up. The hype made us angry.
By 2025, the building blocks were finally here. Foundational models that could reason. Retrieval systems that could traverse graphs. Agent frameworks that could hold state and make decisions.
But something else had happened too: the AI hype cycle had gone off the rails. Everyone was claiming AI would replace software engineers. Founders who had never been paged at 3 AM were building "AI developer" tools. The narrative was all about writing code faster, as if that was the hard part.
We knew better. We'd been the ones patching systems at 2:30 AM. We knew that 83% of an engineer's time isn't spent writing code. It's spent on maintenance, debugging, incidents, and meetings. The industry was solving the wrong problem.
Dave exists because we couldn't wait anymore.
We didn't set out to build another AI tool. We set out to build the agent we wished we had on our team that night in 2017.
Dave is an agentic staffing platform. You don't install a tool. You hire an agent. SRE Dave is the first one: it handles incident response, root cause analysis, alert triage, and runbook automation. It learns from your code, your docs, your logs, and every resolution. And it gets smarter every day.
We built Dave with a simple belief: AI should take the toil, not the craft. Engineers should spend their time on architecture, design, and the creative work that actually moves products forward. Not on being woken up at 3 AM to run the same diagnostic steps for the hundredth time.