Key Intelligence Insight
Fetcherr is not a pricing tool. It is an attempt to replace the decision-making layer of an enterprise with a generative AI model that treats the entire market as a living, simulatable system. The aviation industry adopted it first because it was the most analytically complex and the most technologically neglected. The real bet is that the same model works everywhere.
Founding Story
Fetcherr's origin is methodical in a way most founding stories are not. CEO Roy Cohen and three co-founders, a team carrying roughly 100 years of combined experience, started the company in 2019 as a B2C app. Ten months in, Cohen called investors and pivoted. The B2C model required marketing spend that the economics couldn't justify. The underlying technology could serve any industry. The pivot was to B2B enterprise.
Choosing which enterprise market to enter was equally deliberate. Cohen spent three months evaluating seven industries, producing a SWOT analysis for each. Banking and e-commerce were on the list. The founders had backgrounds in both. The data pointed elsewhere.
Travel won for a structural reason: it was a monopolistic industry controlled by large incumbents that had seen almost no venture capital or private equity investment for five to six decades. Cohen described it as "low hanging fruit." Aviation, specifically, ran on legacy architecture dating to the 1990s. No one had modernized it. No one had tried.
The founding team entered an industry they had no prior experience in, precisely because the incumbents had no real competitive pressure to improve. That gap was the market.
Lean operations were a founding principle, not a reaction to constraint. Fetcherr scaled from four founders to 90 employees globally by keeping costs highly optimized. The main R&D office opened in Poland; the Israeli office sits 25 minutes outside Tel Aviv at 30% of standard market rates. Cohen's stated KPI was explicit: "I never believed in ARR... my KPI is to be cash positive." That orientation shaped every commercial decision the company made.
Product
Fetcherr's core technology is the Large Market Model (LMM), a proprietary generative AI architecture purpose-built to simulate market dynamics at scale. It does not analyze historical transactions. It models entire geographies: politics, commerce, competition, external shocks such as weather events and large-scale entertainment draws, alongside internal variables including website searches, traffic data, looks-to-books conversion rates, competitor price points, load factor positions, historical and shifting average fares, and distribution channel behavior. Cross-referencing all of these previously disconnected variables against an organization's private data, the LMM identifies hidden correlations and predicts market reactions in real time.
The distinction from traditional Revenue Management Systems is structural. Legacy RMS platforms use static fare buckets and rule-based updates. A price change moves in hours or days. The LMM operates continuously, generating millions of forward-looking scenario simulations and executing decisions at a pace and resolution no human team can match.
Three execution engines sit on top of the LMM:
Generative Pricing Engine (GPE): Real-time fare optimization, published automatically across all distribution channels.
Generative Inventory Engine (GIE): Dynamic seat allocation and capacity management, running in tandem with pricing.
Generative Network Engine (GNE): Route planning and schedule optimization. Currently unreleased.
Fetcherr does not position these engines as replacements for existing infrastructure. The mechanism: the system is designed to sit alongside legacy platforms from Amadeus and Sabre, automating the insufficiencies those platforms cannot address. The phrase Cohen uses is "turbo boosting what they cannot do."
Explainability is a deliberate product decision. Every LMM output is auditable. Users can query the system directly in natural language to understand why a price moved on a given day. Fetcherr calls this the "glass box" approach: no black box decision logic, no opaque recommendations. This matters for regulatory compliance and for airline revenue managers who need to trust before they can hand over control.
The result, once trust is established, is a significant shift in how those teams operate. Azul Airlines, one of Latin America's largest carriers, now runs over 70% of its flight network on AI-driven pricing autopilot. The LMM generates more than 3 million fare recommendations annually for Azul, with zero filing errors and 2,000 analyst hours returned to the team each year. Revenue uplift was measurable from day one of deployment.
Market, Competition & Business Performance
Market
The airline industry is Fetcherr's proving ground, not its ceiling. Aviation represents the archetype of the market Fetcherr was built for: complex, high-frequency, analytically demanding, and structurally underserved by modern technology. Airlines manage thousands of flights, hundreds of fare classes, and demand signals that shift by the hour. Traditional RMS platforms cannot keep pace. The humans operating them cannot either.
The broader opportunity is any legacy industry where pricing decisions are complex, frequent, and currently executed with outdated tools. Retail, logistics, hospitality, and financial services each carry versions of the same problem. Fetcherr's expansion thesis rests on the transferability of the LMM architecture: the model that learned to simulate airline markets can, in principle, simulate any market.
The total addressable market for enterprise AI decision intelligence is large and early. Research from MIT indicates 95% of organizations see no tangible results from their AI investments. The gap between AI deployment and AI value creation is where Fetcherr operates.
Competition
The competitive landscape in airline revenue management includes PROS, FLYR, Datalex, Sabre, and Longtail Technologies. Each competes on pricing optimization in some form. None has built a generative market simulation layer of the kind Fetcherr describes.
The more instructive competitive frame is the distinction between the LMM and standard Large Language Models. LLMs generate language. The LMM generates market decisions grounded in real-time economic signals. Fetcherr's founders made this argument directly at Davos in January 2026, positioning the LMM as a categorically different architecture.
The legacy incumbents, Amadeus and Sabre specifically, are not direct competitors. Fetcherr's system is designed to coexist with them. The mechanism: piggybacking on existing infrastructure eliminates the rip-and-replace objection that kills most enterprise AI deals. Airlines do not need to change their core systems to deploy Fetcherr. They add an intelligence layer on top.
The real competitive risk is not from existing players. It is from the possibility that a well-resourced incumbent decides the LMM architecture is worth building or acquiring. That has not happened. The window Fetcherr is moving through is genuine.
There is one competitive pressure worth flagging that has nothing to do with rival vendors. In August 2025, remarks by Delta Air Lines executives triggered regulatory scrutiny and public accusations that Fetcherr's system was executing personalized price discrimination against individual customers. Both Delta and Fetcherr denied it. In December 2025, Chief AI Officer Uri Yerushalmi publicly refuted the claims directly: the LMM does not use personal customer data. It operates exclusively on aggregated public metrics and product-side variables such as seat availability, departure windows, and market conditions. The architecture was designed this way from the start. That design decision now functions as a regulatory moat. Any competitor that has been collecting and pricing against personal data faces a reckoning that Fetcherr has already pre-empted.
Business Model
Fetcherr operates on a cost-plus pricing structure. Cohen's framework is explicit: price everything against actual costs plus a margin, accept only customers who generate profit, and say no to everyone else. Out of more than 150 airlines Fetcherr engaged, the company chose to work only with organizations it describes as "trailblazers": carriers culturally prepared to adopt disruptive AI, not organizations seeking incremental improvements.
The model rejects ARR-maximization. Cohen's KPI is EBITDA and cash positivity. The implication is a smaller, more defensible customer base with stronger unit economics, rather than a large roster of underperforming contracts.
Customer ROI is fast. On a five-year contract, the return on investment is recoverable in under 10 months.
The go-to-market motion is content-led. White papers, case studies, and podcasts function as the primary sales surface. Direct sales is secondary. The logic is that AI adoption at the enterprise level requires significant organizational change management; education precedes conversion.
Traction
Fetcherr's client roster includes Delta Air Lines, Virgin Atlantic, WestJet, Viva Aerobus, and Azul Airlines. These are not pilot agreements. Delta's President Glen Hauenstein described early results as "amazingly favorable" and committed fully. Virgin Atlantic's VP of Pricing, Chris Wilkinson, described the pre-Fetcherr workflow of manually interpreting competitors' pricing, load factors, and channel data as "almost an impossible task." With Fetcherr, Virgin manages 22,000 flights through a system that instantaneously ingests thousands of variables and pushes optimized prices live every day. They run the AI primarily without human constraints, reviewing pricing logic after execution. The result: a material, undeniable profit uplift.
Funding totals more than $150 million. The most recent round, a $42 million Series C led by Salesforce Ventures, closed in September 2025. The stated purpose: global expansion and extension into legacy sectors beyond aviation.
Headcount sits at 219 employees, up more than 46% year-over-year. Engineering grew 59%. Sales grew 60%. HR grew 140%, a signal of structural scaling rather than incremental growth.
Industry recognition includes World's Best Travel Tech Startup in both 2023 and 2024, the 2026 BIG Innovation Award, the Cloud Awards AI Innovation of the Year, the AI Excellence Award, and placement at #4 in 50 Most Promising Israeli Startups and #17 in Sifted's B2B SaaS Rising 100.
Fetcherr appeared at Davos 2026. That is not a vanity metric. It is a signal of where the company positions itself in the conversation about enterprise AI.