Why I Built a Free AI Fuel Finder

134.9. That was the number on the board when I pulled into the station last Wednesday. I did the maths in my head, £74 to fill the tank. I knew there were cheaper stations nearby. I just did not know which ones. So I paid, drove off, and passed a station two miles later showing 129.9.

Five pence a litre. Over a year, that is roughly £50 I did not need to spend.

Every driver in the UK has had that moment at the pump. The tight jaw. The mental arithmetic. The quiet suspicion that you are paying more than you should. You finish filling up, spot a cheaper price down the road, and it is too late. You have already paid.

The problem nobody solved

The UK Government publishes fuel price data for more than 7,600 petrol stations across the country. It is public. It gets updated regularly. Stations report their own prices directly to the feed.

But the data sits in a raw API that no normal person would ever open. It is structured for developers, not drivers.

Price comparison sites exist. Some are decent. But they give you a list of numbers and leave you to figure it out. None of them use AI to tell you what to do. Which station should I pick? Is the drive worth it? What do I save per tank? They hand you a spreadsheet. I wanted to hand people an answer.

So I built something

fuel.richardbatt.com does one thing. You type in your postcode. It searches the government data and returns the 10 cheapest stations near you, with prices per litre, distances, and when each station last updated its numbers. Then an AI reads those results and writes you a plain-English summary. It tells you which station is your best bet, factors in how far you would need to drive, and calculates what you save per fill-up.

No account. No app download. No email required. Open it on your phone at the pump if you want. It works.

Why AI matters here

People hear "AI" and picture corporate boardrooms or research labs. This is not that. This is AI reading a database and telling you, in plain English, where to fill up your car.

The technology behind the fuel finder is the same stuff I deploy inside businesses every week. Reading data. Spotting patterns. Summarising findings so a human does not have to. The difference is the audience. A logistics company uses this kind of AI to process invoices. A recruitment firm uses it to screen CVs. An e-commerce brand uses it to tag and sort customer feedback overnight.

Here, the same thing reads petrol prices and tells you which station saves you £3.50 a tank.

AI as something mundane, useful, and personal. That is the version of AI I care about.

Where I tested it first

I live in Middlesbrough. I fill up at the same stations as you. Fuel prices are climbing again, pushed up by the Iran conflict. That hits harder here than it does in most of the country. Wages are lower. Commutes are longer. Public transport is thin.

A fiver a week saved on fuel is not abstract money here. It is a week's packed lunches. It is the difference between filling up on Monday or waiting until payday Friday.

So Teesside was the first place I pointed it. I ran every TS postcode through the tool before I launched it. The price gaps between stations are bigger than most people think, even within a couple of miles.

How it works (the short version)

The tool pulls live data from the UK Government's fuel price API. That is the same dataset used by official statistics, stations report their prices directly. When you enter your postcode, it searches for stations within your chosen radius (5 to 25 kilometres) and ranks them by price. Then an AI model reads the results and writes a summary in plain English. Which station is cheapest. How far it is. What you save compared to the priciest option nearby.

The data is real. The AI just makes it readable.

What I do for businesses

I have built more than 120 AI projects like this for businesses across 15 industries. Automating weekly reports so a team gets their numbers at 8am without anyone touching a spreadsheet. Triaging support tickets so the urgent ones get flagged in minutes, not hours. Scoring leads so a sales team stops wasting afternoons on dead ends.

The fuel finder took a weekend. Those business projects are what I do Monday to Friday.

If you run a business and want to see what AI can do for you, the AI Ops Vault has playbooks and templates from those 120+ projects. The same tools I use on client sites. If you want a one-to-one assessment of where to start, the AI Revenue Roadmap maps out what to automate and in what order. There is a guarantee: find at least $50K in recoverable revenue or the session is free.

I mention those because people ask. Not because this post is a sales pitch.

Try it

Go to fuel.richardbatt.com. Share it with anyone who drives. It works on any phone, tablet, or laptop, no download needed.

The government publishes fuel prices for every station in the country. That data belongs to all of us. Someone just needed to make it readable.