I’ve spent the last five years working in Boston’s tech scene, but my journey in AI and machine learning has taken me to Glasgow, Toronto, and roles at companies like Amazon and Best Buy.
Along the way, I’ve learned something important: The most powerful AI applications usually come from solving unsolvable problems. Things like clearing out messed up customer data or figuring out why someone bought that laptop instead of four other laptops.
Today, I want to share how we’re using large language models (LLMs) at Best Buy to tackle those challenges. But before we get into the technical details, let me say this clearly: you should not use LLMs just because they are trendy. The business use case has to come first. Always.
A Data Engineer’s Guide to Pipeline Frameworks
If you’re planning for 2026, these are the seven frameworks you really need to care about.
When should you really use LLM for data enrichment?
A question I constantly hear is: should we do this Using LLM For our data problems?
The honest answer is, it depends. And many teams skip that part because generative AI is exciting.
You need the right business use case. If the only tool you have is a hammer, everything starts to look like a nail. That mentality becomes expensive very quickly with an LLM. These models excel at certain tasks, especially when dealing with unstructured data with which traditional ML struggles.
They are great at summarizing text, applying common sense logic, and connecting the dots in disorganized datasets.
For expert advice like this delivered straight to your inbox every other Friday, sign up for a Pro+ membership.
You’ll get 300+ hours of exclusive video content, a complimentary summit ticket, and much more.
So what are you waiting for?
Get Pro+
But they also bring challenges
LLMs can become overwhelmed if you dump too many references into a single prompt. They sometimes ignore instructions that are buried in long prompt templates. And yes, they hallucinate. i really see Maya Less as a bug and more as a side effect of their strengths.
Their ability to extrapolate is what makes them powerful. It just needs railing.
The good news is that costs are falling rapidly. I’ve noticed that token costs have dropped dramatically over the past few years, while model capabilities have improved just as rapidly. This combination opens doors to use cases that were not previously economically realistic.
You also need strong quality assurance processes, clear privacy compliance, and a technical team that is prepared for long-term maintenance. Too many teams focus on the initial launch and forget that these systems require ongoing care.
LLMs are not “set it and forget it” tools. They are like high-maintenance pets. Impressive, useful, but certainly not self-sufficient.
