AI that does one job in your business

Software built for one job, for companies with no technical team.

We work with businesses of every size and sector. One job at a time, at a fixed price, built around the way the work is already done.

Describe the job See selected work

What we build

Not a platform. A tool that removes one task, and nothing else.

Most of what a business needs is too specific for off-the-shelf software and too small for a development agency. That gap is the whole of our work.

Operational tooling

Converters, calculators and checkers that sit between a business's files and whatever consumes them, enforcing the constraints that really apply.

Case and document systems

Portals that replace shared folders and email threads, with the right view for each role and a visible status on every item.

Paperwork pipelines

Extraction and reporting that take handwritten or scanned documents and return structured, consistent output in a house template.

Forecasting algorithms

Demand and sales forecasting built on your own trading history — what to buy, how much, and when — using the same methods large retailers run across tens of thousands of lines.

Pricing and risk models

Prediction where a decision has money attached: where your prices sit against the market, which customers or orders carry risk, and how confident the answer actually is.

Getting the data straight

Prediction needs history, and most businesses have it scattered across systems that disagree. Pulling it together into one reliable record is often the first job, and sometimes valuable enough to be the only one.

A forecast is only as good as the history behind it. Before any model is built we look at what data exists and say plainly whether it supports a useful prediction yet — and if it does not, what would need to be recorded first.

Why now

The gap is opening between businesses that use AI and businesses that read about it.

Large companies are already running on prediction, automation and systems that learn. Smaller ones are adopting at roughly half the rate — not for want of interest, but because nobody has built them anything they can use.

The advantage is no longer reserved

The modelling that used to need a data team and a seven-figure budget now costs a few thousand pounds to put into one business, doing one job. The technology stopped being the barrier some time ago.

Margin, not novelty

Every hour spent retyping, re-checking or re-chasing is margin leaving the business. Removing those hours is where the return comes from, and it compounds quietly every week.

Waiting has a cost

Competitors adopting this now are lowering their cost of doing the same work. That difference does not announce itself; it shows up in pricing, in lead times, and in who can afford to quote.

Selected work

Three jobs, three sectors.

Client names withheld. Each began with a half-day watching the task being done by hand.

Knit chart converter Knitwear manufacturer, Leicestershire

Artwork to machine-ready stitch grid, without colour drift

High-resolution client artwork had to be reduced by hand to a low-resolution bitmap for the knitting machine. Every downscale introduced colours that were not in the design, and each one was found and corrected by eye. The converter reduces the image against the supplied colour legend, so no new colour survives the process, and it refuses any output that exceeds the machine's stitch ceiling. A second converter handles scarves, where the width is capped far lower.

Constraint: 346 × 509 stitches per garment panel; 120 stitches across for scarves
Input: Client artwork plus the colour legend already supplied with it
Status: Working tool, proven on live client designs

Verification portal Employee screening agency

One place to see where every candidate's documents stand

Candidate paperwork moved between HR, the verifying team and the candidate by email, and the answer to "where is that one up to?" meant asking someone. The portal gives each of the three roles its own view of the same case file, with a status on every document and a record of who changed it. Nothing is chased by phone and nothing sits unread in an inbox.

Roles: HR, verifier, candidate — separate access, one case file
Built with: React front end, role-based access, document status tracking
Status: Reviewed and approved by the client; deployment in progress

Field report pipeline Investigation services firm

Handwritten field forms, read once and reported consistently

Investigators return forms completed by hand, mixing Hindi and English with typed sections. Each was read and retyped into a Word report, which took hours and left every investigator's reports looking slightly different. The pipeline reads a photograph of the form, extracts the fields and produces a finished report in the firm's own template.

Input: Mixed handwritten and typed forms, Hindi and English
Output: Structured report in the client's existing Word template
Status: Pipeline built; in testing against the client's form set

Background

Built on prediction work done at national scale.

Before this, the same hands built forecasting and risk models inside large UK businesses — the kind of modelling a small firm is told it cannot afford.

Demand forecasting

Forecasting algorithms for a FTSE-listed retailer, predicting demand across a very large product range and feeding the decisions made on the back of it.

Credit risk modelling

Risk and affordability prediction in UK consumer lending, where a model's output has to be explainable to a regulator as well as accurate.

Competitor price prediction

Modelling competitor pricing from market data, so a lender can see where its own products sit before it moves them.

Approach

You show us the job. You do not write a specification.

Nobody running a business should have to translate their own work into requirements before anyone will listen. That translation is ours to do, and it is most of the value.

Watch the work

Half a day on site, with the person who actually does the task. No forms to fill in beforehand.

One page, one price

Within two working days: what we would build, what it would not do, and a fixed cost.

Build on real files

Four to six weeks, using your own documents and designs, with one review in the middle.

Hand over and keep it running

Training, a written guide, one round of changes, then hosting and small tweaks monthly.

Terms

Fixed price, agreed before anything is built.

No hourly billing and no open-ended discovery phase. If the job turns out to be larger than one tool, we say so and quote the parts separately.

£2,500 – £8,000 one-off

Design, build and hand-over of a single tool. Includes training for the people who will use it and one round of changes after go-live. Half on commission, half on delivery.

From £120 per month

Hosting, backups, security updates and small adjustments. Thirty days' notice to cancel, and the code and data are yours to take with you.

Suitability

We would rather tell you not to buy.

Half of these enquiries are better answered by software that already exists. We will say which, at no cost.

A good fit

Five to a hundred staff, and nobody whose job is IT.
One task takes hours every week and lives in a spreadsheet, a paper form or one person's head.
Standard software has been tried and does not match how the work is really done.
The person who would use the tool every day can be named.

Not a fit

Accounts, payroll, CRM or stock control — good products exist and we will point to them.
An app for the public rather than a tool for your own team.
Anything needing round-the-clock support or handling card payments.
A chatbot for its own sake.

Enquiries

Tell us what still gets done by hand.

A sentence or two is enough to start. We reply within one working day, and the first visit costs nothing.

hello@menderai.uk

Based in

Leicester, England