Building sophisticated AI models means nothing if nobody uses them. This harsh reality hits many data science leaders who focus on technical excellence while missing the business impact. Ahmad Fattahi, Senior Director of Data Science at Cloud Software Group, learned this lesson the hard way when his brilliant predictive model gathered digital dust. His two decades of experience have taught him how to bridge this gap.
Building AI That People Use: When Smart Technology Meets Reality
Fattahi still remembers the sting of building something nobody wanted. Years ago, he created what he calls “a brilliant predictive model that used cutting edge clustering and classification algorithms.” The technical work was solid, the algorithms were impressive, and the results looked great on paper. But there was one small problem. “Nobody wanted to use it,” he admits. That failure taught him something important about AI development. You can have the smartest people working on the most advanced technology, but none of it matters if you’re solving the wrong problem. Since then, Fattahi has focused on “turning data into decisions, leading AI initiatives that improve the top line, reduce costs, and unlock innovation.” His work includes everything from launching Spotfire Copilot to cutting costs and presenting to Fortune 500 executives.
Anchor On Impact, Not Algorithms
Many AI teams get this backwards, according to him. “AI teams, and especially technologists such as myself, are often guilty of starting with models or technology,” he explains. “Making an impact in the real world often starts with the mission.” It sounds simple, but most teams skip this step and jump straight to the fun technical stuff. When his team worked on Spotfire Copilot, they could have started with the latest vector databases or large language models. Instead, they asked a different question. “We need an AI assistant that turns hours of data analysis or process optimization into seconds of insights and action,” Fattahi says. The technical solution came later. When customers see real impact, “they start to engage and share even more with you, creating a virtuous cycle.”
Build As Close To Your Users As Possible
Distance from users kills products. Fattahi learned this from his engineering colleagues, who pointed out “there’s an inverse proportional relationship between the quality of the product and the distance to its users.” The further you get from actual users, the more likely you are to build something nobody wants. For Spotfire Copilot, this meant getting everyone involved from day one. “We had structured working sessions across disciplines with a group of trusted expert users on a regular basis, not just one-off meetings,” he explains. They didn’t wait for traditional handoffs between teams. “We actually started to co-build together with our users.” This approach let them “go from an idea to launch in less than six months.” Better yet, when they launched, they already had early adopters ready to go.
Design Teams for Lasting Success
Enterprise AI products need lots of different skills and people working together. Fattahi figured out that “repeatable success comes from designing teams and processes that mirror the product journey, from innovation to enablement and everything in between.” Many focus only on research and development, but that’s not enough. Fattahi restructured his teams to include “user enablement, customer success, and sales” alongside the technical work. The results were impressive: “Up to 3x team growth, less than 10% attrition due to job satisfaction and smooth operations, and most importantly, a product that your users want to use.” When teams are set up right, everything else becomes easier.
Here’s what He learned after two decades in AI: “Generating value with AI products isn’t about having all the answers on day one.” Success comes from “setting the vision, staying close to your users, and building the right self-correcting processes in the team.” You don’t need to be perfect from the start. You just need to be pointed in the right direction and ready to adjust. Fattahi’s final point hits home for anyone building AI products: “Your product is only as strong as the alignment behind it.” All the technical brilliance in the world won’t save a product if the team, the business, and the users aren’t on the same page. Getting that alignment right is what separates successful AI initiatives from expensive experiments that nobody remembers.
Follow Ahmad Fattahi on LinkedIn to explore more lessons on making AI work in business.
