Egor Karpovich built 12 AI agents in 5 weeks… …and now they’re running parts of his travel company without sleeping.

Watch the video on YouTube: https://youtu.be/IHg6neBdXHw Search can show you ten great trips. Who helps you choose the right one, and negotiate for it? In this episode of the Everything AI in Travel podcast, I speak with Michael Culhane from Mindshare about a proposed agentic marketplace for travel. Michael's career runs from safari lodges in Botswana to travel aggregation and payment systems. He brings an operator's view to a question that has followed him through that work: how can travel businesses connect with the right customer without losing so much value along the way? Got a travel problem AI might solve? Find the AI you need: tell me what's frustrating you and I'll personally introduce you to someone who can help. https://docs.google.com/forms/d/e/1FAIpQLSdVYizRw_UgfX9CpvkeUmkpOeRGFj_iiF0W2KmX_EECIqSUsA/viewform Recorded on 8 October 2026. Michael discusses launch plans for November 2026 and January 2027 as future targets. This episode does not confirm whether those launches happened, or whether the proposed marketplace is available today. We dig into: 🦁 Michael's move from economics into building safari camps and lodges in Botswana 🌍 The role of local communities and skills in a high-value, low-footprint tourism model 📉 Why customer-acquisition cost became his long-running travel-distribution problem 🔎 The difference between finding options and making a useful judgement about them 🧠 Why the same traveller may want something different on a family trip, a business trip or a golf holiday 🤖 His vision for a marketplace where suppliers and travellers each have agents 🎓 Combining specialist travel knowledge with information the traveller gives their own agent 🤝 How supplier-side agents could negotiate offers for particular customer needs 🏨 A hypothetical hotel example of targeted offers rather than public rate cutting 💳 Using existing rewards infrastructure as a route to travel customers 🧾 Where Michael thinks blockchain could help with payments, commissions and reconciliation ❓ My pushback on whether an AI should recommend its own idea of the trip or help fulfil the traveller's existing choices Better search is not the same thing as better judgement. We explore how agents might combine traveller preferences, specialist knowledge and supplier offers, and whether that would help people buy the trip they actually want. Recording-date metadata uses 8 October 2026, the Descript project-created date. Marketplace scale and reach figures are Michael's claims, not independently audited here. Product plans, launch dates and onboarding offers may have changed since. Subscribe for more conversations with the makers, shapers and buyers of AI in travel. Everything AI in Travel newsletter: https://everythingaiintravel.beehiiv.com/subscribe #EverythingAIInTravel #MichaelCulhane #Mindshare

Watch the video version on YouTube: https://youtu.be/W351rvSAq5Y In this episode of the Everything AI in Travel podcast, I speak with Antonio Picozzi, co-founder and CEO of dAirect, about helping hotels get found through AI assistants without handing the customer straight back to an OTA. Got a travel problem AI might solve? Find the AI you need: tell me what's frustrating you and I'll personally introduce you to someone who can help. https://docs.google.com/forms/d/e/1FAIpQLSdVYizRw_UgfX9CpvkeUmkpOeRGFj_iiF0W2KmX_EECIqSUsA/viewform Antonio came into hospitality through a luggage-delivery business after working at Accenture and Amazon. Hotel managers kept asking him about a different problem: how to get more direct bookings. His first answer was a resellable room rate. Now he's working on what happens when AI becomes another layer between the traveller and the hotel. The problem is simple: having a hotel website doesn't mean AI will recommend it. And asking ChatGPT once whether it knows your property isn't much of a measurement strategy. A romantic hotel with a private beach needs to appear in different conversations from a city-centre business hotel. The sources AI uses can differ too. Antonio's approach is to test those specific situations, work on the hotel's digital presence and keep measuring as the answers change. We dig into: 🧳 Why a luggage-delivery startup became Antonio's introduction to hotel distribution 🔄 What resellable room rates taught him about direct bookings, and why building a secondary market proved difficult 🏨 Why fragmentation, trust and loyalty matter as much as giving guests a better rate 🤖 How dAirect tests the answers AI assistants give for a property's particular positioning 📊 Why Antonio says they run 5,000-10,000 prompts per property each month across different markets and assistants 🔍 Why a pre-work baseline and a peer group are more useful than checking whether your visibility went up 🔌 How the work spans the hotel's website, managed digital channels and earned presence ⚙️ Why he wants a product that does the work, rather than another report for a small hotel team to action 💰 The subscription model Antonio describes: around €200-€300 per month, with no booking commission 🛎️ Why tracking the journey through to the booking engine matters more than simply appearing in an AI answer One of the most interesting ideas from the conversation is Antonio's admission that there isn't a fixed secret sauce. His bet is on building a system that keeps testing and learning, rather than selling hotels a formula that assumes AI answers will stand still. For hotels, the commercial question is still the same: who gets the customer, who gets the booking, and what does it cost to acquire them? If AI becomes the place travellers make their shortlist, direct distribution could start well before anyone reaches your website. But visibility alone isn't the win. The booking still matters. More about dAirect: https://dairect.io/ Enjoyed it? Subscribe so you don't miss the next one, and share this episode with someone in travel who'd get something from it. Want the weekly version? The Everything AI in Travel newsletter: https://everythingaiintravel.beehiiv.com/subscribe #TravelAI #EverythingAIinTravel #DirectBookings

In this episode of the **Everything AI in Travel podcast**, I speak with **Emmanuel Annor, co-founder and CTO of Equera.ai**, about using AI to connect some of tourism's smallest and least-digitised businesses to the global travel ecosystem. Emmanuel brings an interesting background to the problem. Before Equera, he built fintech and agritech products used by more than **70,000 farmers across Africa and Asia**, learning how to design technology for remote environments, unreliable connectivity and low-end devices. :chatgpt-content-reference{index="0"} Now he's applying those lessons to travel. The problem Equera is tackling is deceptively simple: **AI can't sell what it can't see.** Some of the most authentic community-based tourism experiences in the world aren't sitting inside reservation systems or OTA databases. They're being managed through **WhatsApp, pieces of paper, phone calls and conversations**. So rather than trying to force these businesses into another extranet, Equera is building technology around the tools they already use. We dig into: 🌍 Why community-based tourism is both economically important and home to some of travel's most memorable experiences 📱 Why Equera abandoned the assumption that suppliers should simply learn to use a portal 📝 The operator who wrote their coffee-tour itinerary **on a piece of paper, photographed it and sent it through WhatsApp** — and how that helped reshape Equera's approach :chatgpt-content-reference{index="1"} 🤖 How AI could turn a WhatsApp message, photograph or even a **voice note in a local dialect** into structured, bookable travel inventory 🔌 Equera's ambition to become an infrastructure layer connecting fragmented offline experiences with OTAs, AI agents and the wider digital travel ecosystem 📅 Why digitisation needs to go beyond simply listing an experience — suppliers also need capabilities such as live availability 🔎 How Equera's status as a **Google Things to Do connectivity partner** can make previously offline community experiences discoverable and bookable through Google :chatgpt-content-reference{index="2"} 📸 The challenge of collecting the photography, video and storytelling required to merchandise these experiences properly ❤️ How that content could also communicate the wider social and economic impact tourism is creating inside communities 🧠 Why Equera is benchmarking different AI models rather than automatically assuming the biggest and most powerful model is the best tool for every job 🤝 And how an emerging world of AI agents and agent-to-agent commerce could create entirely new distribution opportunities for these businesses. There is a bigger issue here that I think is going to become increasingly important. As travellers shift from Google searches and OTA filters toward simply asking an AI **"What should I do?"**, we risk creating a new digital divide. The large, structured and already-digitised tourism businesses are easy for AI to find. The woman running a cooking experience in her village, the family offering a homestay, or the local guide organising experiences through WhatsApp may effectively be **invisible**. As Emmanuel puts it, if we're increasingly depending on AI for travel discovery and booking, we need to make sure the businesses that aren't currently represented in the data don't get left behind. :chatgpt-content-reference{index="3"} That's what makes Equera particularly interesting. It's not simply using AI to make an existing digital travel business more efficient. **It's using AI to help bring an entirely offline layer of tourism into the digital economy.** And potentially giving AI agents access to some of the best stuff in travel in the process. #TravelAI #AI #CommunityBasedTourism #TravelTech #AIAgents #Tourism #Experiences #SustainableTourism #TravelStartups #Africa #Equera

In this episode of the Everything AI in Travel podcast, I speak with Amit Shamni, founder of Bridgify, about building a technology layer for one of travel’s most fragmented categories — tours, activities and experiences — and the journey that ultimately led to Bridgify being acquired by HBX Group.

In this episode of the Everything AI in Travel podcast, I speak with Thomas Bangert, founder of Travecta, about building the connectivity layer he believes multi-day travel has been missing. Thomas has spent 26 years in travel, including more than a decade at lastminute.com, followed by roles with TripAdvisor and Amadeus, consulting work with companies including Sojern and PerfectStay, and time with Exoticca. Travecta is now his fourth startup. The idea for Travecta came from trying to build another travel business and discovering a surprisingly fundamental problem: There was no equivalent of a Hotelbeds, Amadeus or Duffel that they could simply plug into to access multi-day tour inventory. Huge amounts of fantastic local product still live in PDFs, spreadsheets, emails, WhatsApp conversations — or simply inside the heads of local DMCs. Travecta wants to turn all of that into structured, instantly bookable and globally distributable inventory. We dig into: 🌍 Why so much multi-day tour inventory still isn't bookable online 🔌 How Travecta wants to become the infrastructure connecting local DMCs with OTAs, airlines, cruise lines and other global distributors 🤖 How AI can ingest a PDF or other unstructured tour information and transform it into standardized, distributable inventory ⚡ The challenge of moving traditional DMCs from "on request" bookings to genuine instant confirmation 🧑‍🤝‍🧑 Why Thomas believes human-to-human trust remains critical — even as more of the infrastructure becomes automated 🔗 Why Travecta has been built MCP-native and is preparing for agent-to-agent travel commerce 💬 How future AI agents could potentially do far more than search — negotiating changes to itineraries, duration, inclusions and price 💳 Why the distributor remains merchant of record while Travecta essentially becomes the invisible "rails" connecting both sides 📄 How automated contracting could eliminate the enormous workload involved in hundreds of DMCs individually contracting with hundreds of distributors 🧩 Why personalization doesn't necessarily mean rebuilding an itinerary from scratch — Travecta can offer a "skeleton tour" with optional components layered around it 🏨 Why Thomas believes AI-created dynamic packages and locally crafted DMC tours will coexist, rather than one replacing the other 📦 And the fascinating longer-term opportunity to both bundle and unbundle multi-day tours — potentially allowing distributors to access individual accommodation or experience components from DMC inventory. Travecta has already begun onboarding DMCs, with hundreds more going through its process, while the distribution side is being prepared for launch. Thomas also makes a big prediction: He believes consumer behaviour is heading toward agent-to-agent travel booking, and that within the next couple of years a significant proportion of bookings could happen this way. If that happens, the companies controlling the structured, queryable and bookable inventory underneath those agents become very interesting indeed. #AI #TravelAI #TravelTech #AIAgents #MCP #ToursAndActivities #MultiDayTours #DMC #TravelStartups #TourismTechnology #TravelDistribution