George Ralak and his friends had looked forward to touring Watamu for the longest time. They had made the plans and saved the money. Only on arrival did the question catch up with them. “Now that we are here, what do we do?”
They scrolled social media for places to visit and restaurants to try, and eventually found a tour guide. Having someone follow them everywhere was not the getaway they had imagined.
“We kept wishing that there was one platform where we could go and get everything we needed to know about Watamu, where to go, what to do, and who to hang out with, but there wasn’t one such platform,” George says.
The same thing happened on later trips. Then the friends realised they had the skills to do something about it.
“We all work in tech. Four of us are software developers, and I am a product designer, so we just decided to build the platform ourselves.”
That was the beginning of Tukai, a Kenyan-built website and mobile app with AI features. Users pick interests from coffee and fitness to spas and outdoor activities, then receive recommendations. They can save bucket lists, much as people save places on Instagram. The platform lists more than 3,000 places, each identified and verified by a dedicated team.
Tukai is also a marketplace for people who organise experiences, from boat rides to trivia nights.
“Someone looking for something to do on a Saturday evening could discover a quiz night organised by Nairobi Games and book a ticket through the platform,” George says.
To protect users, Tukai vets hosts before they can list and holds ticket money until the experience has taken place. Hosts pay Sh130 a month to list, and Tukai earns a commission or processing fee on ticket sales. Browsing and the current AI recommendations are free for travellers.
The founders plan to explore paid subscriptions for more advanced AI features, such as personalised itineraries and help with bookings.
The first version reached a small group of users in March 2026, with a second following in September. Tukai now has about 5,000 sign-ups and up to 8,000 weekly visitors, with traffic rising from Thursday through the weekend.
Manjunath Gowda, founder of Twiga.ai, during the Magical Kenya Travel Expo at Uhuru Gardens in October 2026.
Photo credit: Michael Ochieng’ | Nation Media Group
George is careful not to celebrate too early. “The uptake is quite nice, but we are not there yet. The only proof that this is working is when these numbers reflect repeat users.”
“Event organisers want to know that you have people coming to your platform, but the people want the guarantee that your platform has quality experiences. It is difficult to get one person without the other, so you have to try and get them both at the same time.”
Working with government agencies has been a challenge too, especially the duration it takes to get permits as well as fees charged.
“This is a local product. We are doing it here because we love our country, and it would really be nice if it was easier for local creators to work with the government.”
George believes Tukai’s data could show how visitors spend their time and reveal gaps in Kenya’s tourism offering. The company wants to expand into other African markets and become a travel assistant that helps people plan from start to finish.
“We want to get to a point where travellers will reach for Tukai every time they think of places to visit or things to do,” he says. Businesses, he adds, “can just come to Tukai, add their details, and have our AI handle everything else, including bookings and payments.”
While Tukai serves the traveller who has arrived, Twiga.ai serves the people who bring them here. Its owner, Manjunath Gowda, believes a trip begins the moment a traveller reaches out to an operator.
“The experience starts from the time they touch the operator,” he explains. Speed, he says, can decide everything. “When they reach out to you, if you don’t get back to them at the right time, within a given time, they will go to some other country or other operator.”
He pictures a family planning a December trip and sending the same request to a Kenyan operator and an Indian one. The Kenyan company replies in half an hour and the Indian company the next day. “Where will they go? They go to Kenya.”
A fast reply is hard to give because of hotel contracts. Rates shift with the season and the room, and children carry their own pricing.
Twiga.ai takes that burden off their shoulders. Its AI reads the contract and turns the rules into an equation, and a separate maths engine does the counting. “Our mathematical engine and the AI completely calculate to the exact cent.” What once took days can now take minutes. “In five or 10 minutes, it can give a quote to a customer.”
The platform also manages itineraries, customer records, vouchers and accounts, with each operator’s own markups applied. “It is the operating system for safari and adventure travel operators,” Gowda says.
The company is almost three years old, and the team spent about a year and a half building the software before selling a single subscription.
“The AI part was very tough,” Gowda recalls. They kept going, and sales began last October and November. Twiga.ai now has 20 customers in Eastern Africa. In its first year, it had three or four users. “Today, we have more than 100 users.” He puts the growth at about 300 per cent, with each user paying about Sh5,100 a month.
Twiga.ai does not do marketing, so it cannot bring them more leads. What it can change is how many leads become bookings. “Assume they (operators) got 100 leads. They were probably converting 2 or 3. Now they are converting about 20 through the app.”
He adds that the software can steer travellers toward places they have never heard of. “Why does everybody come to Masai Mara?
Because that is the only place they know. Why not Samburu? Why not Tsavo?” If that works, visitors spread out, and so does their money.
He is honest about the challenges, such as the possibility of AI getting facts wrong. His own background is in cybersecurity, and they have someone who checks everything the AI produces. “We have a manual oversight for all the AI. So that AI hallucination will not happen.”
Then comes the question many in the industry ask: Will AI take jobs? Gowda says tourism is already feeling the shift. “A lot of jobs may be replaced.” He has no tidy answer for families who depend on them. “I don’t have a ready-made answer for that.”
Yet he sees work no machine can do. “It cannot drive a safari vehicle.” If more visitors reach quieter parks, he says, those jobs could grow.
Next, he wants an AI agent that answers travellers in real time, and a small language model built only for safari travel in East Africa.
