Company profile · 3 min read
Prox turns equipment catalogs and rules into governed product specialists
Prox helps equipment manufacturers connect product knowledge, calculators, compatibility rules and quote workflows across buyer and channel surfaces.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Prox is a technical sales assistant for equipment manufacturers. The fit is a manufacturer with a complicated catalog, rules and quote process—not a company looking for a generic website chatbot.
What it does
Prox turns product knowledge into a governed product specialist. Its public material describes catalogs, manuals, calculators, compatibility rules, accessories and quote logic connected behind a website, channel portal, QR code or messaging workflow. The buyer starts with a job or constraint—rather than a model number—and Prox is supposed to ask for the missing information, apply approved rules and move the result into a quote, dealer, service or parts path. Prox homepage YC profile
The company’s public example uses concrete equipment decisions: a buyer describes a concrete-wall job and constraints, then the product recommends a panel mix with an alternative and cites the source material. That is a better product definition than “AI product search.” It is governed selection and handoff.
Prox’s documentation also exposes a remote connector for Claude. It describes read-only catalog and knowledge-tree tools plus versioned, reversible edits to product knowledge. Prox MCP documentation
Why I’d look closer
Equipment manufacturers often have product knowledge spread across manuals, spreadsheets, applications engineers, dealers and support teams. A governed layer can make that expertise available at more entry points without forcing a buyer to know the SKU first.
The buyer gate is the approval model. Prox says experts test real questions before launch and approve the experience. That is the right shape for a product where a wrong recommendation can create a return, a service failure or a bad quote.
What could make it the wrong choice
Implementation is the work. The manufacturer has to provide product data, rules, calculators, compatibility constraints and the source material experts trust. A polished answer is not enough; the team needs a way to keep the product graph current and review corrections.
Pricing is not published in the checked sources. The public docs describe authenticated workspace access and an edit path, but they are not a third-party security assessment. A buyer should ask about permissions, audit logs, data retention, environment separation and what a dealer or customer can actually change.
My editorial take
I would shortlist Prox for an equipment manufacturer where wrong-fit recommendations and slow quote handoffs are expensive. I would not start with a broad catalog import. Start with one product family, one calculator or compatibility rule, and a reviewable quote path.
The interesting product decision is whether the company has enough governed product knowledge to become a sales surface. If the answer is no, Prox is not a magic layer; it is a reason to clean up the knowledge underneath.
Cohort context
Prox is listed in Fall 2025. In our 2026-09-18 directory snapshot, 90 of 146 listed companies in that cohort have YC’s primary industry label B2B (61.6%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:15:17.400Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Not observed in this response |
| Docs/developer link | Observed |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗