Why I Pivoted from Big-Firm Legal Tech to Something Much Simpler

For the past two years, I’ve been building legal tech for large firms - mostly around eDiscovery and document review. It’s been a journey.

The tech worked well, but I underestimated how incredibly slow that market moves: long sales cycles, endless approvals, multi-year vendor contracts.

Now I’m rethinking direction and trying to understand where smaller firms and individual attorneys actually struggle the most.

One thing that keeps coming up in conversations is client intake and case assessment - attorneys spending a decent amount of time on initial conversations with clients, collecting basic info to qualify cases.

And the more I talked to attorneys, the more I realized how fundamental this issue is. Intake isn’t glamorous, but it’s the gateway to every case. If it’s slow, inconsistent, or poorly structured, everything downstream becomes harder. A few attorneys told me they lose strong cases simply because they don’t respond fast enough, or because clients drop off during long forms and intake calls.

At some point, it clicked for me: this wasn’t just an operational annoyance - it was a real bottleneck that directly affects revenue and client experience. And unlike enterprise legal workflows, it was a problem that smaller firms wanted solved now, not next year.

That’s essentially how the idea for Retriever was born: a simple, configurable AI assistant that handles the initial client conversation, gathers facts, reviews documents, and helps attorneys quickly understand whether a case is a fit. Nothing over-engineered, no heavy onboarding — just solving one painful, universal, day-to-day problem.

Still early, still learning, but so far it feels like the right direction.