Hello OmniTalk Fans!
One of the biggest questions facing retailers right now is how to actually put AI to work. The technology is getting smarter by the day, but there’s still one major problem: AI doesn’t know your business.
This week, Chris Walton sat down with Arber Sejdiji, Founder & CEO of Zenline AI, for another edition of 5 Insightful Minutes, and the conversation offered a surprisingly simple way to think about the challenge. In fact, Arber gave us an analogy that Chris says he’ll never forget: an AI agent is like your newest employee on day one.
Here’s what stood out from the conversation:
Your AI Employee Doesn’t Know Your Business Yet
Imagine hiring an incredibly smart, highly educated employee. On day one, they may have all the skills you could want, but they don’t know your company’s terminology, products, processes, workflows, or history.
That’s essentially how Arber thinks about AI agents.
The missing ingredient isn’t intelligence. It’s context.
As an employee gains experience, they accumulate the knowledge needed to make better decisions. Arber argues that retailers need to do the same thing with AI by capturing years of merchandising knowledge and turning it into what he calls a “company brain.”
Context Engineering Could Be the Key
Arber describes this process as context engineering, or giving an AI agent the organizational knowledge it needs to actually do its job.
His background helps explain why he approaches the problem this way. After studying engineering at ETH Zurich and spending three years at Boston Consulting Group working on technology, AI, and business transformation, Arber saw firsthand that technology is only part of the equation.
As he put it, digital transformation is roughly 80% people and processes and 20% technology.
The lesson for retailers is pretty simple: even the most powerful AI models won’t improve an organization if they don’t understand how that organization actually works.
Why Start With Merchandise Planning?
Zenline AI is taking on one of retail’s hardest problems: merchandise planning.
That might sound like a slightly insane place to start, but Arber thinks that’s exactly the point.
Retailers are constantly trying to answer two fundamental questions: What do my shoppers want, and what are they willing to buy? Arber argues that these questions are still often answered in surprisingly manual and unanalytical ways.
AI changes what’s possible.
Zenline can combine traditional numerical and tabular data with unstructured information from sources like TikTok, social media, and the broader web to help retailers understand what shoppers actually want and turn those insights into recommendations for pricing and merchandising.
The Hardest Problems Might Finally Be Solvable
Arber acknowledges that merchandise planning is difficult. But that’s also why Zenline chose it.
The technology has changed enough that AI can now work with much more unstructured information without the weeks or months of preprocessing that traditional machine learning often required.
And that creates an opportunity to tackle problems retailers have struggled with for years.
Arber says Zenline can identify gaps in a retailer’s assortment by comparing what shoppers are looking for with what the retailer and its competitors are offering. The system can even look internationally for emerging trends and younger brands that may signal where shopper demand is heading next.
Speed Might Be the Biggest ROI
Perhaps the most compelling part of the conversation was just how quickly these insights can become actionable.
According to Arber, retailers can begin identifying actual assortment gaps within the first couple of hours. For retailers with fast enough internal processes, those insights can then make their way onto shelves within weeks.
And that speed matters.
Merchandising decisions don’t happen in isolation. Once a retailer decides it wants a product, it still has to figure out inventory, shelf facings, placement, and other operational decisions. The faster retailers can identify what shoppers want, the more time they have to actually act on it.
The Future of Retail AI Is About Knowing Your Business
The biggest takeaway from our conversation with Arber wasn’t necessarily that AI can replace merchandising teams.
It was that AI could become dramatically more useful once it understands the organization it is working for.
Retailers have spent years accumulating data, processes, and institutional knowledge. The next challenge may be figuring out how to make all of that knowledge usable by the AI agents they’re now deploying.
Because maybe the future of retail AI isn’t about finding a smarter employee.
Maybe it’s about finally giving the employee you already have enough context to do the job.
Until next time,
The Omni Talk Team
You can listen to or watch my full conversation with Arber Sejdiji of Zenline AI wherever you get your podcasts.
Apple Podcasts | Spotify | Soundcloud | Amazon Music | YouTube
Be careful out there,
Chris and the Omni Talk Team
Music by hooksounds.com
Sponsored Content
Omni Talk® is the retail blog for retailers, written by retailers. Chris Walton founded Omni Talk® in 2017 and have quickly turned it into one of the fastest growing blogs in retail.