Skip to content

Intelligence / AI chatbot

AI chatbot development, trained on what your business actually knows

Trained on your documents and product data, citing its sources, escalating what it cannot handle. Live on your site in ten days for $1,997.

Fixed price
$1,997
Ships in
10 days
Quote in
48 hours
Start this build

Code transfers to you. Deposit back if we miss the date.

Most businesses answer the same forty questions forever. Shipping times, pricing, what is included, whether you work with their industry, how to cancel. Somebody is paid to type those answers, and the ones asked at 11pm go unanswered until morning.

This is a chatbot trained on your actual material: help docs, policies, product pages, past tickets, PDFs, whatever holds the answers. It responds in your voice, links to the source, and hands off to a human when the question is outside what it knows.

Ten days, $1,997, deployed on your site and connected to wherever your team already works.

Best for

  • Support teams answering the same questions daily
  • E-commerce stores fielding shipping and returns questions
  • Service businesses qualifying inbound enquiries
  • Any site where the answers exist but nobody can find them

Built with

  • OpenAI
  • Anthropic Claude
  • pgvector
  • Supabase
  • Next.js
  • Vercel

What's included

Everything in the $1,997 scope

Written out in full, because a fixed price only means something when the scope is specific.

01

Knowledge base ingestion

We take your website, help centre, PDFs, spreadsheets, Notion pages, and historic support threads, then clean and chunk them for retrieval. Messy sources are normal and handled. This is most of the work that makes a bot accurate.

02

Retrieval-grounded answering

Every response is built from retrieved passages and cites the document it came from. That is the difference between a chatbot customers trust and one that confidently invents a returns policy.

03

Tone and persona calibration

The bot writes the way your business writes. We tune it against real examples of your support replies, including what it should refuse to discuss.

04

Human escalation

Below a confidence threshold, or on trigger phrases like cancel and refund, the conversation routes to email, Slack, or your CRM with the full transcript attached.

05

Evaluation set and accuracy report

Fifty real questions with correct answers, scored before launch. You see measured accuracy rather than a vibe, and you keep the harness to re-test after any content change.

06

Site widget and deployment

A styled chat widget matched to your brand, installed on your site. WordPress, Shopify, Webflow, Next.js, or a plain script tag.

What you receive

The concrete list handed over on ship day.

  • A trained chatbot deployed on your website
  • Branded chat widget matched to your design
  • Vector knowledge base built from your content
  • Source citations on every answer
  • Human escalation to email, Slack, or CRM
  • Conversation logging and an analytics view
  • Accuracy report against a 50-question evaluation set
  • An admin route to update content without a developer
  • Thirty days of tuning support after launch

What this tier does not cover

Stated up front. Anything here can be quoted as an add-on.

  • Taking actions in other systems, which is the AI agent tier
  • Voice or phone channel
  • Multilingual support beyond one language, priced per language
  • Ongoing content maintenance after the tuning window
  • Model API usage, billed to your own account

On-time guarantee

Miss the timeline on a fixed-price build and your deposit comes back. No negotiation, no invoicing games.

How it runs

From brief to live in 10 days

The same four phases on every build. No discovery retainer, no phase two.

  1. 0148 hours

    Scope lock

    You pick a tier or send a brief. We come back with a written scope, every screen, every integration, a fixed price, and a ship date. Nothing starts until you approve it, and nothing outside it is assumed.

  2. 02Day 1

    Asset handoff

    A single onboarding form collects logins, brand assets, and access. One call if the project needs it, none if it does not. We select the stack and start the same week.

  3. 03The bulk of it

    Build in the open

    Senior delivery from day one, no junior handoff. You get a walkthrough video at every functional milestone and a live preview URL you can open at any point, so there is no reveal at the end.

  4. 04Launch + 30 days

    Ship and hand over

    It goes live on your infrastructure, in your accounts, with the repository transferred. Thirty days of support covers the bugs and questions that only appear once real users arrive.

FAQ

AI chatbot questions

Answered here so you can decide without a sales call.

How much content do I need for it to be useful?

Enough to cover the questions you actually get. In practice that is often a help centre and two or three policy documents. If the answers do not exist anywhere yet, we will tell you during scoping: writing them is the prerequisite, and no model fixes a missing answer.

What does it cost to run each month?

Typically $20 to $80 in model usage for a few thousand conversations, billed by OpenAI or Anthropic to your own account. We give you a projection at your expected volume during scoping and build caching in to keep it low.

Can it book appointments or look up an order?

Reading from a system, such as checking order status, can be added for $500. Taking actions: issuing refunds, rescheduling, updating records, is agent territory and belongs in the AI agent tier.

How do we update it when our pricing changes?

Through an admin route we hand over: re-sync a source and the knowledge base rebuilds itself. No developer, no redeploy, no call to us.

What happens when it does not know something?

It says so and escalates, with the full transcript, to wherever your team works. It does not guess. Escalation rate is a metric we report, and a high one usually points at a content gap worth filling.

Can it handle more than one language?

Yes. The base tier covers one language; each additional language is $300 and adds a day, mostly for evaluating quality rather than for the translation itself.

Other AI app services

AI agent

An agent that finishes a real process across your tools, not one that answers questions about it.

Price
$4,997
Timeline
14 days

RAG system

Retrieval over your whole document library, with citations and access control.

Price
$5,997
Timeline
14 days

Start a ai chatbot build

$1,997, shipped in 10 days. Send a brief and we will confirm scope and a start date within 48 hours.

No retainer, no hourly billing. Quote in 48 hours.