A product of OrangeFin

Answer every product question, without adding headcount.

A product assistant that answers from a manufacturer's approved manuals and specification sheets, at any hour.

Try it at productchatiq.com ↗

The ProductChat IQ home page, headed "A specialist product assistant for every job on your site".

What a question actually costs

Technical teams lose time to repeat questions about ratings, fit, warranty and installation. Customers who ask after hours wait for an answer or look elsewhere.

A generic chatbot can fill a missing fact with a confident guess. For technical products, that is not an acceptable answer.

What changes for the business

Approved answers arrive immediately, specialists keep their time for complex work, and recorded questions reveal where product information needs improvement.

A product built around trusted answers

The product page puts the customer problem first: approved knowledge is turned into immediate answers, with the source still visible and a human handoff when the material is silent.

The current ProductChat IQ product page explaining how approved product knowledge becomes sourced customer answers.

You decide what it is allowed to say

Manuals, specification sheets, warranty terms and product pages go into a controlled library. Only published sources can reach a live answer, and each assistant reads only the material assigned to it.

The ProductChat IQ document library, listing specification sheets and warranty terms with their extraction and publication status.

Separate assistants for separate jobs

The assistants page shows how each experience can be given its own approved material, role and escalation path instead of relying on one generic bot.

The current ProductChat IQ assistants page explaining controlled assistants for different products and customer questions.
Current live-site capture.

A clear path from source material to a live assistant

The implementation path is made visible before a buyer commits: prepare approved material, configure the assistant, add it to the site and improve the experience from real questions.

The current ProductChat IQ how-it-works page outlining the path from approved source material to a deployed support assistant.

Why you can put it in front of customers

The product is designed around one risk: a confident wrong answer.

Numbers are checked against the document

Every number is checked against the source material retrieved for that answer.

It says when it does not know

Unsupported questions are handed to a person instead of filled with a guess.

Every answer names its source

The source document stays attached so customers and staff can verify the answer.

Built to be run, not just demonstrated

The controls behind the interface make it safe to operate for more than one company.

Separation enforced by the database

Documents and conversations are isolated for each company at the database level.

A bot cannot run up your bill

Session, address and daily limits control automated use and cost.

Deletion that cannot go wrong quietly

Conversation retention follows a customer-defined schedule with safe defaults.

The brand

OrangeFin built the name, identity and interface rules so an answer, caution and refusal are visibly different.

  • Product Ink #17343B Headlines, body text, the wordmark
  • Deep Teal #0B5D61 Primary actions and dark fields
  • Signal Aqua #31C7B7 Accent and selected states. Never body text on white
  • Source Mint #DDF4EF Source panels and supported answers
  • Support Slate #60767B Secondary copy and metadata
  • Warm White #F8FAF8 The page background

Read the ProductChat IQ brand guidelines PDF, 10 pages, 72 KB One panel on page 9 shows a source citation the shipped widget does not display. Corrected on this page.

Built with

  • Cloudflare Workers
  • Cloudflare Pages
  • Workers AI
  • Supabase Postgres with pgvector
  • Supabase Auth
  • Private storage buckets
  • TypeScript
From $12,500 CAD An application of this shape, a multi-tenant application with authentication, a relational database, and per-tenant isolation, is scoped under our full engagement model. Most engagements are considerably smaller than this one, and start with a single tool solving a single customer decision.

What this would be in your business

Connected to your data sheets, it becomes an always-available product support assistant.

Do your customers keep asking your team the same questions?

Start a conversation