Wondering if AI could help your business? The right software system certainly can.

A business approached us recently and asked us if we could help them, because they needed an AI system. We said great, let’s talk about it to find out more. As we delved into their current system, its limitations and what they wanted to achieve, we realised that they didn’t actually need AI. What they needed was a software system designed to solve their particular challenges.

There’s an interesting phenomenon here. Widespread awareness of AI, particularly tools like ChatGPT, has raised expectations about what computer systems can do for you – and we love that. Too many businesses have spent too many years accepting third rate software, and it’s holding them back. We want businesses to get the tools that truly make a difference to how they work and their results. AI is making people think again about the power of great software – even if AI isn’t always the answer. And we welcome this.

So do you need AI?

You may well need help from AI. Or you may need a smart system that uses other kinds of logic, algorithms or processes to deliver the results you need. For example, AI can be used for demand forecasting based on pattern recognition – but so can software based on mathematical formulae developed by theory of production experts.

In the end it doesn’t matter whether the solution uses AI or not (we use both AI and other kinds of software in our systems). What matters is getting software specifically designed for the needs of your business, not an off-the-shelf package that continually limits what you’re able to do.

Let’s look at an example.

Anticipating fluctuations in customer demand

Imagine your business has hundreds of products, most with limited shelf lives, and customers who depend on getting the right product when they need it. If you can’t supply a customer with a product they need, you’re providing poor service, losing sales, and risk losing a customer. But if you make certain you’re never short of any product, you’ll be constantly throwing away stock that hasn’t been sold. It’s a difficult dilemma faced by thousands of businesses in the food industry and far beyond it.

One such business, providing quality wholesale food to the catering and hospitality industries, used to run their business using paperwork and instinct – summer’s coming, so let’s increase strawberry stocks. But as the business grew, the variables grew too – what difference did bank holidays make? Wimbledon? School holidays? It became too complex for a human being to predict accurately across hundreds of lines of fresh food.

We built a software system that did many things for them. One was to predict demand for hundreds of products across the year, taking into account dozens of variables. We used some smart maths developed for stock forecasting, inputted their own data, tested and tuned the results, and provided an accurate forecasting module. The system also monitors stock levels, sell-by dates and market prices so it can alert the business when it’s running low, automatically place orders, reduce prices on over-stocked items to increase demand, and much more. In other words, it does that vital forecasting work, but it integrates this into a wider system to deliver much broader benefits. 

It’s all about data

In this case we didn’t need to use an AI. Our forecasts use only data from within the systems and algorithms developed for the task, and it gives highly accurate and specific results. In other cases, where the demand fluctuations are affected by vast numbers of inputs, an AI might be a better solution – although AI has its limitations, including hallucinations, opaque processes, and time-consuming training.

Either way, the key is having the data. That’s why our systems always collect data, whether or not it’s immediately useful. The data might be about your customer enquiries – and later you might want to correlate this with marketing activities or seasonal changes. The data might be on outputs from various processes through your workflow – and later you might want to look closely at this to see where unnoticed production bottlenecks are occurring.

So, what do you need?

What you really need is a software system that solves your real problems. Those problems might be fluctuating demand, slow manual processes or lack of insight into productivity issues. Which is why we don’t make assumptions about what your real problems are. We take a considered, step-by-step approach to understand what’s really going on under the bonnet.

We’ll unpack why you’re asking for the solution you want. What is important to you? What outcome are you trying to achieve? Is the problem you’ve identified really what’s stopping you achieving that outcome? What data do you have?

In this way we can understand what the best solution will be. We embrace new tech but we always stick to our founding principles: be inquisitive, understand the real problems and desired outcomes, deliver a specific solution.

Business systems can provide wonderful, game-changing solutions – with or without AI. But only if you find a partner who understands how businesses work, not just how software works. That’s why we welcome AI. Because it’s taught people to expect more.