Quanfey Inc. — Toronto, Canada
Quanfey helps logistics, service, and workflow-heavy teams replace scattered inboxes and manual follow-up with AI-assisted operational systems — SaaS products, Atlassian Forge apps, AI assistants, and automation we build and run ourselves. Canadian, founder-led. Start with one working session.
Status, risk, and next action — pulled live from the tools you already run.
Assistants and agents that read, extract, and decide — and show exactly how sure they are.
Intake, routing, SLAs, and dashboards — so nothing waits to be chased.
Our systems run live — in our own products and in operations we've built for others.
Most teams don't have a work problem. They have a visibility problem. Status lives in inboxes, spreadsheets, and ticket fields — and nobody fully trusts any of it.
01 — What we build
The layer that fixes it: AI that reads your documents and shows you exactly how sure it is, platforms that pull one operational truth out of the tools you already run, and automation that routes work before anyone has to ask.
Assistants and agents that read, extract, classify, and decide — with confidence you can see on every result.
See how we solve it →Command layers that turn tickets, emails, and attachments into one live view of status, risk, and next action.
See how we solve it →Intake, routing, SLAs, and dashboards, engineered end-to-end so work moves without being chased.
See how we solve it →02 — One example
QEI reads an invoice and tells you the vendor, the dates, the totals — and exactly how confident it is on each field. Below the threshold, a human sees it before it counts.
That's our AI philosophy in one screen: automation you can act on, because it tells you when to check.
03 — Flagship in the field
Our strongest proof isn't a pitch — it's a courier operation running on a system we built end to end. Intake, workflows, live queues, SLA early warning, dashboards, and governance, delivered as one command layer for the people dispatching real trucks.
The same architecture is now becoming product — OpsFlowHub and our AI-assisted command apps grew out of this build.
04 — The deliverables
Everything this page promises is running in software right now — a live marketplace, an AI document engine, and a workflow command console. The same architecture we deliver to clients, proven on ourselves first.
A live marketplace for Canada's military families.
A live, production SaaS marketplace connecting Canadian Armed Forces families with base-adjacent housing, pet services, and community resources. Built end-to-end: Stripe billing, Supabase/Postgres with row-level security, map-based search — independently operated, zero government affiliation claimed.
Next.js 14 · React 18 · Supabase (Postgres + RLS) · Stripe · Mapbox + Leaflet · Vercel + Cloudflare
AI confidence you can see.
Upload or scan an invoice or bank statement. Claude reads it — vendor, dates, totals, tax — and tells you exactly how confident it is on every field. Anything under 80% is flagged for a human before it counts.
Next.js 15 · React 19 · Convex · Clerk · Anthropic Claude · Inngest
One operational truth per shipment, built on the Jira you already run.
Turn scattered Jira tickets, emails, and attachments into one deterministic view of every shipment: status, evidence, SLA risk, and next action — no black-box AI, fully auditable logic, built on Atlassian Forge.
Atlassian Forge · React 19 · TypeScript · Vite · Jira REST API v3 · JSM Cloud · ScriptRunner (Groovy)
05 — How our systems are built
AI results carry a confidence score, and anything uncertain is routed to a human before it counts.
Money, status, and compliance run on auditable rules — never on a model you have to take on faith.
The same architecture we implement for clients runs our own products in production.
Quanfey Inc. is a Canadian AI systems and workflow platform company that builds AI assistants, AI agents, automation, and operational software that turn scattered tickets, emails, and documents into one clear operational truth. We work with teams anywhere from our base in Canada.
Both. We design and implement AI and workflow systems for operational teams, and we build our own software products using the same architecture.
No. We use AI where uncertainty can be shown and reviewed, and auditable rules where money, status, and compliance are on the line.
With a working session: you bring the workflow that hurts, and you leave with a clear read on how to fix it — typically within one business day of reaching out.
Next
Bring the workflow that hurts. Leave with an opinion on how to fix it — ours, in plain language, in one call.
