Think about the last time you had a question about your finances. It was probably very recent, and likely went unanswered. Maybe you wanted to know how much you spend in a month, or if you could have gotten something for cheaper, or what your K.Points were actually worth. Chances are, you either dug through the app trying to find the answer, asked someone, or just let the question go.
That gap between having a financial product and actually understanding it is exactly what K.ai was built to close.
Today, we're proud to introduce K.ai, Klivvr's in-app AI assistant and the first of its kind in the region. K.ai is natively embedded in the Klivvr experience, and built around one simple idea: your money should make sense to you.
Why We Built It
At Klivvr, we've always believed that financial tools should empower people, not overwhelm them. We've worked hard to build products that are clear, human, and genuinely useful: a credit limit with a vast network of merchants, a Family Plan that makes household tracking at your fingertips, a loyalty program that gives you flexibility on how to redeem your points.
But products alone aren't enough. Even the best-designed financial app can leave you with questions. And for too long, those questions have required you to call customer support, dig through help centers, or make decisions based on instinct. K.ai was our way of changing that.
We don't say this lightly: K.ai is the first AI assistant natively integrated into a fintech product in the region. A purpose-built, contextually aware assistant trained on the actual features, flows, and terminology of the product.
K.ai doesn't give vague answers pulled from the internet, it sees what you see, meaning that it understands your credit limit, your due payments, your K.Points balance, what plan you’re on, and your family spending. It can walk you through available offers, repayment options, explain how point redemption works, and help you understand what's happening in your insights, all while speaking your language (literally). Financial wellbeing is a day-to-day practice, and K.ai is here to make it accessible.
Built to Feel
For Hossam Saieed, the designer behind K.ai, the brief was never just visual. “We wanted K.ai to feel like an evolution of Klivvr rather than a different product living inside the app.” That philosophy shaped every decision, starting with the entry point. The floating button meant immediate discoverability and persistent accessibility without disrupting the existing app architecture, positioning K.ai as a smart layer across the entire product, not a feature you have to go looking for.
The visual language followed the same logic. The gradients and glow effects weren't decorative, they were functional, emphasizing “moments where the AI feels active or thinking, subtly reinforcing the idea that K.ai is alive within the experience.” The bottom sheet format kept things lightweight and contextual, so users stay in their flow rather than being pulled out of it.
The hardest challenge, Hossam says, was balance: “Simplifying the experience enough for first-time AI users while still making the product feel powerful and advanced.” AI inside fintech can easily feel overwhelming or overly technical, so the experience was designed to feel trustworthy, human-centered, and assistive rather than controlling. Looking back, what he's most proud of goes beyond aesthetics: “How much strategic thinking went into the experience: from user trust and interaction psychology to shaping how AI should behave inside fintech products in the region.”
Built to Know
For Kirolos Rezkalla, our Data Scientist, building K.ai came with a different kind of pressure. “It has higher stakes than a normal chatbot, it's people's money.” That meant accuracy had to come first, always. K.ai never answers from memory. It reads your real account data and verified product information before responding, and if it doesn't have an answer, it says so instead of guessing. “If K.ai says it, it’s verified.”
Getting smarter over time works differently here too, and deliberately so. “It doesn't change on its own, and that's on purpose.” Automated checks on real conversations surface gaps, and the team fixes issues at the source, continuously keeping up as people find new ways to ask about money. The hardest part, Kirolos says, was understanding why K.ai answered the way it did. “With this kind of AI, you can't just read the code to know.” The solution was leaning into observability, tracing every step behind a single answer. That same approach carried through to Egyptian Arabic, where users mix languages, drop letters, and use slang that no rigid system could anticipate.
What he's most proud of is simple: “You can ask about your own money, in your own words, and get an answer that's actually relevant to you.” But it goes further than that. Tell K.ai what you want to buy and from where, and it works out an installment plan against your actual credit limit — real numbers, grounded, verified, and holdable. That is what makes K.ai more than a chatbot. It's a financial tool that actually knows you.
We’re just getting started
K.ai will keep growing. We're committed to expanding its knowledge, sharpening its ability to understand how people really talk about money, and making it more personalized over time. The version you're meeting today is the foundation.
So go ahead — ask K.ai anything.
