Inside the Tehnopol AI Accelerator: Narfe wants to turn your favourite travel video into your next trip

You’ve been there: scrolling through a travel video, spotting a string of places that look incredible, and then facing the tedious part – figuring out what those places actually are, where they’re located, and how to fit them into an itinerary that makes sense. Estonia-based startup Narfe World thinks that gap between inspiration and planning shouldn’t exist.

“Travel inspiration and travel planning have become disconnected,” says founder Abdul Salam Mohammed. “People discover amazing places through travel creators and videos, but then have to jump between videos, Google, maps, booking sites and other tools to actually build the trip. Narfe is trying to connect those two worlds – turning travel inspiration into a practical, personalised itinerary that you can actually use.”

From a personal frustration to a product

The idea didn’t come from a grand theory about the future of travel; it came from an ordinary problem. “You watch a great travel video, see several places you want to visit, and then the real work begins: finding those places again, checking where they are, figuring out what fits together and eventually building an itinerary,” Mohammed explained. That led the team to a simple question: why couldn’t the video itself become the starting point for the trip?

That question became Narfe – a platform that takes knowledge from travel creators, runs it through AI and location intelligence, and turns it into something a traveller can actually act on.

Three different backgrounds, one product

Narfe was built by a three-person founding team whose backgrounds barely overlap, which the founders see as an advantage rather than a coincidence.

Mohammed, the founder, spent more than 12 years in market intelligence, strategy and consulting across Europe and India, and focuses on the commercial and product side of the business. CTO Afroz Ahmed Khan brings over 20 years of software engineering and technology leadership, translating ambitious product ideas into working systems. CMO Jack Flowers has more than a decade in research and insights, including UX research, and keeps the team grounded in what users are actually saying rather than what the team assumes they want.

“That combination of market intelligence, engineering and customer research has shaped how we’ve built Narfe,” Mohammed said.

Not trying to out-chatbot ChatGPT

With general-purpose AI tools increasingly used for travel advice, an obvious question is why anyone would need a dedicated app for it. Mohammed doesn’t see Narfe as competing with ChatGPT or Claude at their own game.

“If you ask a general AI, ‘give me a five-day itinerary for Japan,’ it can produce a perfectly reasonable answer,” he said. “But Narfe starts somewhere different: real travel content and the people who created it.” The platform identifies actual places and experiences mentioned in creator videos, maps them geographically, cross-checks information across multiple creators, and builds an itinerary from that evidence – with the sources visible, rather than hidden inside a black box.

“So for us, AI is the technology enabling the experience,” he added. “The product is Travel Intelligence built around trusted human travel content.”

Behind a simple-looking upload flow is a multi-step process: Narfe scans a video’s content and metadata for travel-relevant information – specific places, restaurants, attractions, activities – then has to resolve those mentions into real-world locations, since creators often use informal names or skip addresses entirely. Once locations are geo-mapped, the system combines them with the traveller’s own preferences – destination, dates, interests, pace – to build a day-by-day plan. The team has also extended this into what they call Creator Travel Intelligence, cross-referencing multiple creators’ videos to surface overlapping recommendations. “The difficult part isn’t generating nice travel text,” Mohammed noted. “It’s turning messy, unstructured creator content into reliable, geographically useful information and then into a trip that makes sense in the real world.”

Betting against subscriptions

Rather than charging a monthly fee, Narfe operates on a pay-per-plan basis – the first itinerary is free, and subsequent ones cost a small fee. The decision came out of user research: most people don’t plan trips often enough to justify another recurring subscription. “The simpler proposition is: when you need Narfe to plan a trip, you pay for that plan,” Mohammed said. “There’s less commitment for the traveller and the value they’re paying for is very clear.”

But pay-per-plan isn’t meant to be the whole business. The team sees a much bigger opportunity in the commercial intent that sits right behind a finished itinerary – bookings for accommodation, experiences and flights. Longer term, Narfe is also being built as an embeddable “Travel Intelligence” layer that airlines, online travel agencies and other travel businesses could integrate directly into their own products, instead of losing customers to third-party research and planning tools elsewhere. “So long term we see a combination of consumer revenue, travel-commerce revenue and B2B/B2B2C technology partnerships,” Mohammed explained.

What the accelerator actually changed

Narfe joined the Tehnopol AI Accelerator after already having a working MVP – the team wasn’t looking for help building a prototype, but for people who would stress-test the business itself. “We wanted people who would challenge the business around the technology – customer validation, positioning, go-to-market, pricing and whether what we’re building actually solves a problem people care enough about,” Mohammed said.

One example stands out. The team initially assumed revenue sharing would be the strongest hook for getting travel creators to join Narfe’s partner program. Mentor conversations and direct feedback from creators told a different story: many cared more about visibility, brand collaborations and sponsorship opportunities than a straight cut of revenue. That insight is reshaping the Creator Partner Program: Narfe is now building dedicated creator profiles and actively promoting creators through its own channels, rather than treating a revenue split as the only value on offer. “That sounds like a relatively small change,” Mohammed said, “but it came from listening rather than defending our original model.”

The five-year ambition

It’s clear that being a “leading AI travel planner” doesn’t just mean amassing users who type destinations into a chatbot. The bigger goal is for Narfe to become the intelligence layer sitting between travel inspiration and travel booking – whether someone discovers a destination through a creator, through Narfe itself, or eventually while already browsing an airline’s app. “Narfe should be able to understand what that person wants, bring together credible creator knowledge, turn it into a personalised and geographically practical trip, and connect that plan with the services needed to actually travel,” Mohammed said.

The company’s first major consumer markets are the United States and the United Kingdom, while it builds out partnerships and its creator network internationally, with Europe and other outbound travel markets seen as natural next steps.

“Five years from now, success would mean Narfe isn’t known simply as another AI itinerary generator. We want travellers and travel companies to think of Narfe as the technology that turns trusted travel inspiration into trips people can actually take,” Mohammed concluded.

Tehnopol AI Accelerator is accepting applications until September 25, 2026. Read more here!