Kortex Labs Book a call

What does your AI project actually cost?

Six questions. A real range, based on what production builds actually cost, not a number designed to get you on a sales call. Nothing is sent anywhere; this runs in your browser.

What are you building?

The shape of the thing matters more than the features.

How messy is the data?

This is the single biggest driver of cost, and the one most quotes ignore.

How many people will use it?

Does it need to connect to your existing systems?

Who maintains it after launch?

How exposed is it if it gets an answer wrong?

Determines how much evaluation and guardrail work is real work, not optional.

About 4 weeks

$8,000 – $11,000

A production assistant over your own documents, with citations and evals.

    These are real build ranges, not a lead-capture number. If your answers put you under $6,000, the honest advice is usually that you don't need a custom build yet, and I'll tell you that on the call.

    How these numbers are built

    Most agency quotes price the demo and discover the rest later. These ranges price the parts that actually consume the weeks:

    • Data preparation — usually 30–50% of an assistant build: scanned files, three versions of a policy, nothing marked current.
    • Evaluations — a set of real questions with verified answers, run on every change. Skip this and you cannot tell whether a change helped.
    • Integration — every legacy system adds a week you did not plan for.
    • Handover — repos, docs and a runbook someone else can deploy from. Without it you own a dependency, not an asset.

    More detail in the full cost breakdown.