In a world of instant gratification powered by smartphones, the car buying experience can often be firmly stuck in a pre-iPhone era. At least that was Tarek Kabrit’s experience when he was walking down a Beirut street in early 2015 and spotted a Mini Cooper he liked. Curious as to what a car like that would sell for, he thought of the app Shazam, which can identify songs just by “listening” to a clip, informing the listener of the track title and artist’s name. Mr Kabrit wanted to be able to snap a photo of the Mini to get its exact make and model year and what he could expect to pay on the used market. But no such service existed for looking up cars; as the saying goes, necessity is the mother of invention. A typical car buyer spends 17 hours over the course of three or four months researching vehicles, pricing and where to buy, according to Edmonds, a used car site in the United States. Mr Kabrit helped to create the app he wanted for himself, becoming the co-founder and chief executive of Seez, a meta search engine that can condense months of research and turn the process into a single search that takes a matter of seconds. Seez can translate a photo of a car snapped on a smartphone or a bungled keyword search such as "nissan Petrol 2016" and clean up the cacophony of car adverts available online to provide vehicle specs and the fair market value, as well as what’s available to purchase nearby from dealers and individuals looking to re-sell, and a forecast of how much the car will depreciate over time. “People are spending way too much time and I want to streamline the process,” Mr Kabrit says, adding that another shortcoming of the traditional process of buying a car is “there is no room for negotiation in the classifieds experience”. Enter Seez’s AI-powered chatbot, called Cesar. When a user searches on Seez, they can opt-in to having Cesar reach out to sellers on WhatsApp to begin the process of negotiating the price down. Cesar has so far successfully negotiated around 13 per cent of car sales on Seez without any human assistance. Along the way, the chatbot has improved his negotiation skills, including cutting the word "bro" from its extensive vocabulary - women on the other end of the negotiation were, understandably, not fans. When Mr Kabrit moved from Beirut to Dubai in 2015, it was to join Yahsat - Mubadala Investment Company's space company - as a merger and acquisitions director. Before that, he had worked for Abraaj, which was the largest buyout fund in the Middle East and North Africa until it collapsed last year following a fallout with investors over the alleged use of their money in a $1 billion healthcare fund. Mr Kabrit recalls that "culturally there were some issues" at Abraaj but working from the Beirut office he was less affected, adding that he is grateful to have left long before the firm began to unravel. His work experience as a fund manager at Abraaj did, however, provide strong introductions to venture capital funds and during his time there he was tapped to become a venture partner at Wamda Capital. Before long he was eager to be on the other side of the table - pitching his own start-up to potential investors instead of being on the financing side. Mr Kabrit and his nephew, Andrew Kabrit, put their heads together about the idea for a Shazam for cars. Andrew had a side business while growing up fixing iPhone screens and was pursuing a business degree in Copenhagen with a focus on AI. He was eager to get involved. In September 2015, while still working full-time for Yahsat, Mr Kabrit established Seez in Dubai Silicon Oasis. By February 2016, the first version of the Seez app launched with image recognition and pricing data. That year, the pair also rounded up a small team of coders from Denmark - mostly contacts of Andrew's - to work on improving the app. A few Danes with coding chops moved into a penthouse in Dubai, half paid for by Mr Kabrit and the other half out of their new Seez salaries, and they worked in the living room - with partitions of cardboard set up between their work stations - coding into the small hours of the morning. In May 2017, after putting $420,000 of his own savings into the company, Mr Kabrit secured $1.8 million in seed funding. A few months later, the team launched the new version of the app called Seez R - premiering its chatbot negotiater, Cesar. Then Mr Kabrit left his full-time job at Yahsat. To use a hackneyed driving metaphor, the road to creating the world’s first price-negotiating, car-buying chatbot has been a long one. Mr Kabrit and his team of 10 - now scattered across the world working from Dubai, Lyon, Beirut and Denmark - are looking ahead to expanding into new countries such as South Africa and translating Seez for its Arabic audience in Saudi Arabia, its newest market. Yes. We launched Seez when people were frustrated by the lengthy process and time that it took to find a car to buy online. People wanted a new way to find the best deals since the regional economies had slowed down. When we started building the Seez app in 2015, artificial intelligence had become advanced enough to develop and launch some key features of Seez, such as our car image recognition and price comparison. Chat bots were also just taking off so we had the tools and knowledge necessary to develop Cesar, the world’s first car negotiation chat bot. Our technology is extremely complex and it’s important that it all works in sync. As soon as you snap a picture of a car you like, our AI recognises the make and model, estimates the fair price and its depreciation, and then searches across 16 websites to display where it is advertised. Our AI then starts contacting and negotiating the price with hundreds of sellers on your behalf. All of this happens within milliseconds. Over the years, we have collected comprehensive car data to train our AI. We have recently published our annual UAE report which details the $32 billion-dollar car buying market, including top listings by make and origin, market supply, least depreciation models, and supply and demand. We are proud to work with many of the big automotive companies, providing them with data and AI modules. I like start-ups which solve a basic problem that has existed forever, to make our world more efficient. TaskRabbit, for example, organised people to run errands for one another. Airbnb uses real estate more efficiently and they are very design focused, which I appreciate too. I have learnt many new things throughout this journey. Following a 17-year career working in private equity and venture capital, I thought I knew a lot about how start-ups and businesses work - but I was very wrong. I’ve learnt to take step back from the business and question what I am doing. When you do this, you give yourself an opportunity to fine tune your value proposition and avoid going in a suboptimal direction. It’s easy for company founders to get caught up in the day-to-day activities while running a business. Suddenly, a year has passed and perhaps you had the wrong business model or you were addressing the wrong problem to solve.