Prox

AI product specialists for equipment manufacturers

3.5Score#248 of 999 agents#7 of 20 in Sales assistant

Why it ranks #248

Prox, an AI sales assistant from YC F25, ranks #248 of 999 AI agents and #7 of 20 in Sales assistant, with a score of 3.5 of 10. Data as of 3 Oct 2026.

The score adds four parts, each 0 to 10, by weight. Every point comes from public evidence. How scoring works ·

Proof 3 × 30% 0.9
Scale 3 × 30% 0.9
Momentum 3 × 25% 0.8
Autonomy 6 × 15% 0.9
= 3.5

Proof
Do customers use it? 30% of the score
3
Claims named customers.
- +3
  Named customers
  “We deployed our first agents at ShipBob (unicorn 3PL handling 100+ million packages per year) to automate carrier claims for lost-in-transit packages.”
  Launch YC post · 11 Nov 2025
Not found: revenue or growth (+4), a customer result (+2), a customer count (+2), work done at scale (+2)
Scale
Is there a real company behind it? 30% of the score
3
7 people.
- +3
  7 people on its YC page5 to 9 people
Not found: funding round in its news (+3)
Momentum
Is it shipping and growing now? 25% of the score
3
Shows hiring.
- +3
  Hiring: 5 open rolesFounding Engineer at Prox, Founding GTM Engineer, Executive Assistant & OperationsYC page
Not found: launch YC post in the last 6 months (+3), new traction claim in the last 6 months (+2), funding news in the last year (+2), a batch in the last year (it is YC Fall 2025) (+1)
Autonomy
How much of the job does it do on its own? 15% of the score
6
Supervised agent: takes actions in other systems; a person approves key steps.
- +6
  Level 2 of 4: Supervised agentTakes actions in other systems; a person approves key steps.
What the model read
- Acts in other systems: 71% likely
- Finishes whole tasks: 31% likely
- Calls itself an agent: 74% likely
Level 2: Takes actions in other systems; a person approves key steps.

About Prox

Prox turns product expertise into revenue for equipment manufacturers.

Equipment buyers rarely begin with a model number. They begin with a job, an installed system, a constraint, or a problem. The knowledge needed to qualify, recommend, configure, quote, repair, replace, or upgrade the product is scattered across manuals, catalogs, spreadsheets, calculators, product managers, applications engineers, dealer processes, support teams, and legacy systems. The result is abandoned demand, slow qualification, wrong-fit recommendations, quote delays, weak dealer follow-up, missed parts and service revenue, and dependence on a few experienced people.

Prox builds AI product specialists for equipment manufacturers. A product specialist uses the manufacturer's approved product knowledge, expert rules, calculators, and connected systems to move customers and frontline teams from a complex product question to the right purchase, quote, dealer, parts, service, replacement, or upgrade path.

How it works:

1. Understand what the person is trying to accomplish. 2. Resolve the product, application, and operating context already available. 3. Ask only for the missing information that changes the result. 4. Apply approved sources, expert rules, and deterministic tools (calculators, selectors, configurators, comparisons). 5. Produce a supported, cited recommendation, configuration, procedure, or decision artifact. 6. Move the person into the correct cart, quote, RFQ, BOM, dealer, sales, parts, service, replacement, upgrade, or expert action with the context preserved.

One product specialist, many governed skills: product discovery and comparison, sizing and configuration, competitive defense, quote preparation, dealer enablement, installation and commissioning, troubleshooting, parts identification, and approved service, replacement, and upgrade paths. Manufacturers deploy those skills to customers, dealers, sales teams, applications engineers, support teams, or field service. Existing CRM, CPQ, commerce, and support systems stay in place.

Support and service are revenue-bearing surfaces. The specialist resolves the customer's problem first, then surfaces the approved part, service, replacement, or upgrade path only when the evidence supports it.