Kismet wants to bring hotels to conversational search

What is your 30-second pitch to investors?

Kismet is a managed AI-commerce cloud that puts hotels inside conversational search. Built on an open-source, travel-specific Model Context Protocol service, we turn scattered
hotel data into AI-ready rates, content and booking actions—then serve it from the edge in under 300 milliseconds, with zero operational burden or lock-in.

When a potential guest asks ChatGPT for “the best Vegas hotel next Tuesday in their price range,” Kismet makes it possible for hotels to surface and secure the booking on the spot. Think of Kismet as Vercel for travel suppliers—same usage-based model,
same open-source software community flywheel.

Describe both the business and technology aspects of your startup.

Kismet sells a managed AI-commerce cloud to travel suppliers, starting with hotels of every size. The underlying open-source travel MCP service is free to self-host, so any property can experiment without budget friction. We monetize on two complementary
layers:

 

Revenue layer	What the hotel pays for	Why it matters
Edge usage fees	A tiny per-call charge each time an AI agent (ChatGPT, Google SGE, OTA bot, brand chatbot) hits the hotel’s MCP endpoint for live rates, content or a booking action	Aligns cost with real AI traffic; future-proof because we keep the edge API compliant as new standards (NLWeb, MCP updates, OTA schemas) emerge—no reintegration hassles
Core Compute Subscription (our AI Data Infrastructure Plan)	A fixed monthly fee that covers the heavy lifting: ETL pipelines, vector database storage, continuous data re-indexing, PCI/GDPR compliance and 24/7 SLA	Removes the intensely technical work—crawling PDFs, menus, FAQs, rates; structuring them into AI-readable vectors; hosting them on GCP—so hotels are discoverable and bookable without hiring data engineers

This pay-as-you-grow structure eliminates up-front capital expedinture, delivers SaaS-style gross margins for us and perfectly aligns our revenue with the volume of AI-driven demand each hotel actually sees.

The result: A property can drop a one-line script, be discoverable in conversational search within 24 hours and scale to millions of AI queries without hiring infrastructure engineers.

Give us your SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis of the company.

Strengths:

• The future is here, and we’ve been building servers and databases for AI to operate since before MCP was announced.
• Kismet’s managed cloud takes the technical lift away from hotels that would find it difficult to do on their own.
• Our open-source travel MCP removes lock-in risk and fosters a developer community to support hotels of all sizes.

Weaknesses:

• We still need to educate the customer–this is SEO and SEM on steroids, and hotels are getting caught flat-footed against the OTAs.
• We’ve taken the risk on hosting costs as data usage intensifies—obviously we think this is worth the first-mover advantage, but it’s a challenge.

Opportunities:

• Rapid growth of AI trip-planning (ChatGPT, Google AI Overviews) means this is all happening to hotels whether they like it or not, and the good news is they are structurally advantaged versus the OTAs if they invest in this now.
• In a few years…moving beyond hotels: our destination-graph data model combined with the open-source travel MCP lets us add spas, tours, restaurants—new potential customers for Kismet—and partnership opportunities for our hotels.

Threats:

• Large incumbents (e.g., Lighthouse, OTAs) could push proprietary MCP pipes. Mitigant: this would be a terrible choice for hotels to rely on a closed MCP service since the application layer to operate it will be limited to what the incumbent would
  provide.
• Lethargy: The hotels have lost this fight through inaction before. Mitigant: aggregation (i.e., an OTA marketplace model) is a worse business model in AI distribution than in the traditional Web 1.0 framework travel has been stuck in since 1996.

What are the travel pain points you are trying to alleviate from both the customer and industry perspectives?

For the guest/traveler:

• “Where should I stay?” is generic: Today’s AI agents can’t access deep hotel info, so answers feel generic.
• Fragmented booking flow: Bouncing from chat page to OTA page is annoying and adds clicks.
• No real personalization: Consumer AI right now lacks structured data on amenities, vibe, and experiences that matter to the individual guest.

For the hotelier/industry:

• Invisible in AI search: Unstructured websites and PDFs leave properties off the conversational map.
• Heavy OTA dependence: 15-25% commission bites margins; no control over end-to-end experience
• High technical barrier: ETL, vector databases, compliance and edge delivery are beyond most hotel IT know-how, let alone budgets.

Now that the product is built, what’s your strategy for customer acquisition?

• We solve today’s problem: MCP is the future, but AI can’t book—yet. We help not just with surfacing rates for booking but in driving traffic with significantly more important marketing content in a hotel’s own voice. Discovery comes first!
• Open-source MCP repo + developer software development kits: Seed a community that builds plug-ins and spreads Kismet inside tech-savvy chains; hotels can try without risking lock-in.
• Usage-aligned pricing : Hotels start on a low-cost Core tier and pay only as AI traffic grows; zero capex lowers friction.

Tell us what process you’ve gone through to establish a genuine need for your company and the size of the addressable market.

We’ve been talking to hotels about their AI needs since we started the company at the end of 2023. The hard-won insight was that data structure and silos were the limiters to AI adoption. So, we’ve been building data transformation and storage pipelines
and “MCP servers” since before “MCP” was even a term.

But to put it in hotel terms, the good news is that you don’t have to do AI. Chatbots on your website aren’t the future. No guest wants a virtual concierge. But you do have to let their
personal AI interact with the content and data you already produce: your social media content, press coverage and website. We make that possible with no new work for you. 

How and when will you make money?

We already do.

What are the backgrounds and previous achievements of the founding team?

Jason Cincotta is a former hotelier with data expertise.

Nathaniel Sena has been in social media and tech for a decade.

How have you addressed diversity and inclusion within your business?

We’re a pre-seed company today. We commit, as part of our seed round, to earmark funds for paid contributor grants specifically aimed at open-source contributors from under-represented developer communities.

What’s been the most difficult part of founding the business so far?

The speed at which everything in AI has developed. I thought we had five to 10 years before commerce was happening. We had two.

Generally, travel startups face a fairly tough time making an impact—so why are you going to be one of lucky ones?

Most travel startups ask hotels to change what they’re doing. We’re asking them to lean into what already makes them special, and let us take care of the tech.

A year from now, what state do you think your startup will be in?

The obvious choice for hotels that want to go direct-to-guest.

What is your endgame? (Going public, acquisition, growing and staying private, etc.)

We’re agnostic. Travel is a big total addressable market, but what gets us out of bed in the morning is a world where a hotel’s best guest finds them easily and transacts directly. As long as we can help make that happen, we’ll consider it a success.