research · 2026-07-27 · 3 min read
GPT Is Not the Right Tool for Every Data Problem
Why large tabular forecasting tasks often need LightGBM or CatBoost instead of a GPT model, and what to do before choosing an AI approach.
People still reach for GPT when they have a business dataset and want an answer fast. I get why. It feels like the shortest path from a spreadsheet to something that sounds useful. But if the task is forecasting sales, product demand, or another problem built on large tabular data, a GPT model is often the wrong tool. It was built for text and context, not for rows of numeric history.
Compare tools for the task
For structured sales or inventory data, LightGBM and CatBoost are candidates worth testing alongside simple statistical baselines. They are not automatic winners: the choice depends on the available history, the forecast period, and performance on data the model has not seen. Language models can still help explain results or let people ask questions about them.
Prepare the data first
A mistake I see a lot is people assuming they can send the dataset to ChatGPT and get a real forecast back. GPT can help in smaller ways. It can summarize a few months, write Python code, or give you a first-pass analysis if the data is already reasonably clean. But if the source files are messy, split across formats, or badly structured, the work starts earlier: preprocessing, aggregation, and feature design.
That part matters more than people want it to. If you want a useful prediction, you have to decide what the model should actually learn from the business. Seasonality. ABC analysis. Product attributes. Inventory behavior. Whatever matters in that category. The model is only as good as the inputs you shape for it, and weak inputs usually give you a nice-looking mistake.
A project example and its limits
We ran into this recently in a Zenit-Auto project on sales prediction and product demand. The goal was simple to say and annoying to do well: predict demand with low error so the business can order the right products in the right quantities. That is not a “just ask GPT” problem. It is a modeling problem, a data problem, and a judgment problem.
Our reported test result was a WAPE of 43%. That number alone does not establish forecast quality or a business improvement. To judge it, we would need the forecast period, product coverage, test data, and a comparison with the existing approach. The practical lesson is to compare methods on the same decision and data before choosing one.
Planning an AI automation initiative? Read the AI automation POC guide.