About Poth Labs

We're building persistent customer knowledge teams can actually inspect

One system that brings together what your customers have told you—and stays honest about what the evidence cannot answer.

01

The problem we kept seeing

Companies already capture useful evidence in sales calls, support tickets, CRM notes, surveys, and Slack threads—scattered across different tools and owners. Decisions then lean on whoever remembers loudest, and the same customer problem can be rediscovered every quarter.

02

What we built

Poth is a persistent knowledge layer for customer data. It connects the sources you already have and organizes accounts, people, features, and issues into an evidence map where every relationship traces back to the source behind it.

03

Why it's different

A general model can produce a fluent answer even when the supplied data is incomplete. Poth is built around the evidence: answers cite their sources, confidence uses retrieval signals, and contradictions and coverage gaps are surfaced instead of smoothed over.

Founding team

A technical team obsessed with answers you can trust

Matthew Wong portrait

Cofounder & CEO

Matthew Wong

Cofounder & CEO @ Poth Labs (YC S26). Ex-Palantir. Stanford.

LinkedIn
Mojmir Horvath portrait

Cofounder & CTO

Mojmir Horvath

Cofounder & CTO @ Poth Labs (YC S26). Ex-Tietoevry, Phillips Academy and HTL-Spengergasse.

LinkedIn

Principles

What we believe

Evidence over vibes

An answer without evidence is only a hypothesis. Poth is designed to return citations to retrieved quotes, tickets, metrics, or notes so readers can inspect what supports a result.

Honest about uncertainty

"We don't know yet" is a valid answer — and often the most useful one, because it tells you exactly which customers to go talk to next.

Data privacy first

Your data is yours. We process it, analyze it, and help you understand it — but we never train models on it or sell it.