Lowe’s puts project expertise into every hand
OpenAI ● Covered by 2 sources
Lowe's teamed up with OpenAI to build two AI assistants: one for shoppers, one for store staff. The goal is basically cloning a helpful hardware-store expert and putting them in everyone's pocket.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Home improvement retail has a staffing problem that's been true forever: the guy who actually knows how to re-tile a bathroom or wire a ceiling fan doesn't work every shift, at every store. Lowe's answer is to build that expertise into software instead of hoping he's on the clock when you walk in.
The company partnered with OpenAI on two tools. Mylow is customer-facing, meant to walk shoppers through a project from a vague idea — say, wanting a better patio — through the actual list of materials, tools, and steps needed to pull it off. Mylow Companion is the internal version, built for the roughly 300,000 people who work the floor at Lowe's stores, giving them a way to quickly pull up product specs, project guidance, and troubleshooting help instead of relying purely on memory or tribal knowledge passed down from senior associates.
What's notable here isn't the novelty of a retail chatbot — plenty of retailers have bolted one on. It's the specificity of the use case. Home improvement projects are messy: a customer might not know the right fastener for a deck versus a fence, or which primer actually matters for a bathroom with humidity problems. Generic AI assistants stumble on that kind of granular, physical-world knowledge. Lowe's is betting that grounding a model in its own product catalog and project data produces something genuinely more useful than a search bar or a static how-to article.
There's also a workforce angle that's easy to miss. Associate turnover in big-box retail is high, and training someone to give confident advice on a countertop installation takes time most stores don't have. If Mylow Companion actually shortens that ramp-up period, the payoff isn't just customer satisfaction — it's fewer costly mistakes and returns, and less pressure on the handful of veteran employees who currently carry that expertise around in their heads.
The real test, as always, is what happens at the shelf. Assistants that sound smart in a demo can still send someone home with the wrong grout or the wrong voltage rating. Lowe's is essentially wagering that AI can compress years of hands-on retail knowledge into a tool anyone can use on day one — and that's a much harder problem than swapping in a chatbot for FAQs.
My take — AI-written commentary, not fact-checked reporting
This is one of the more sensible enterprise AI rollouts I've seen lately because it's narrow and grounded in real inventory data rather than trying to be everything to everyone. Retail is exactly where AI should be proving itself first — boring, high-volume, low-stakes-if-wrong use cases — instead of the constant push toward AGI hype. If it actually cuts down on associates giving bad advice about which anchor bolt to buy, that's a bigger win for AI's reputation than another chatbot benchmark score.
Read more about this at: OpenAI