On 12 October 2026, PROS introduced PROS Kairos, which it describes as an agentic airline retailing platform that brings its pricing science, airline data and AI into a single workspace. The announcement is light on deployment detail and heavy on architecture and positioning. That makes it more useful as a signal of where airline commercial technology is heading than as evidence that it has arrived.
What PROS announced
Kairos gives analysts one role-based workspace in place of several separate applications, with a conversational interface that lets them query data and trigger platform capabilities in natural language. PROS says the workspace is tailored to each user’s role and permissions and saves preferences across sessions. The release singles out Product Management tools, Product Catalog and Stock Keeper as the areas where that role-based tailoring applies, and it does not say what is available today.
Underneath sits a data layer built on the Databricks Data + AI Platform, which PROS says it is adopting as the core foundation for Kairos analytics and AI work. Above it, a PROS Agent Orchestrator is designed to coordinate specialised agents through open-source Model Context Protocol connections. Analysts interact with a single platform agent, while the orchestrator is designed to coordinate the specialised agents behind it. PROS adds that it plans to extend agents across offer optimisation, dynamic pricing, shopping, group sales optimisation and offer marketing.
The launch follows PROS’s February 2026 decision to focus exclusively on travel retailing, and the company is pitching Kairos as something airlines can adopt in pieces on top of the PROS products they already run.
The positioning matters more than the interface
The most revealing passage in the release is the CEO’s framing. Sunil John says airlines need to add capabilities quickly without having to choose between a fully custom build and a multi-year commitment to a single closed suite. PROS’s own language elsewhere in the release stresses commercial independence and advancing toward Modern Airline Retailing on the airline’s own terms.
TDN reads that as a deliberate counter to the suite model, in which an airline’s commercial stack is bought largely from one vendor. The release does not name a competitor, so this is interpretation rather than something PROS has stated. But the language on open architecture, adopt-what-you-need modularity and independence points clearly in that direction, at a time when concentration among the major airline technology providers is already a live industry concern.
Why MCP is the detail to watch
Using the Model Context Protocol as the integration layer between agents is the most technically specific claim in the announcement. MCP is an open standard, so in principle it could let an airline connect specialised agents, including ones PROS did not build, into the same workflow without bespoke integration work for each. The release does not say that, however. It refers to specialised agents and open-source MCP connections but never states whether third-party agents are supported. That answer decides whether the architecture is open to outside agents or only to PROS’s own.
The release says the orchestrator is “designed to” coordinate agents this way, which describes intent rather than a working capability. It does not say whether any third-party agents connect today, what governance sits around those connections, or how an airline controls what an agent is allowed to do inside a pricing or inventory workflow. Those questions carry real weight in a function where a wrong action has a direct revenue cost.
The Databricks choice cuts both ways
Anchoring Kairos on Databricks gives PROS an established data and AI platform on which to build its own airline retailing capabilities, and it lets the company argue that airlines can connect market, behavioural and operational data without migrating to a single platform. Shiv Trisal of Databricks makes the case that AI agents are starved for context when they only see siloed transactional systems.
The trade-off is that an airline pursuing independence from a closed commercial suite is now being asked to accept a dependency on a third-party data platform at the core of the architecture. That may be an acceptable bargain, particularly for carriers already on Databricks. But the release does not address what it means for airlines on other data stacks, or how the claim that intelligence developed once can be reused “regardless of where data resides” works in practice.
What is verified and what is not
Everything above comes from PROS’s own announcement as published on its website, and TDN has not independently verified the capabilities described. The release names no airline customers for Kairos and gives no pilot or go-live dates. It offers no performance figures, no revenue or efficiency claims, and no description of what an agent actually does on a given task beyond giving analysts natural-language access to information. Much of the agentic functionality is stated as intent, with phrases such as “plans to extend” and “intended to help.” The one fully concrete element is the named set of initial tools and the Databricks relationship.
There is also little in the announcement specific to emerging markets. Whether a platform of this kind suits carriers with thinner data infrastructure and smaller commercial teams is an open question, and arguably the most relevant one for airlines across Africa and the Middle East.
TDN’s view
Kairos is best understood as a statement of intent from a company that has narrowed its focus to airline retailing and wants to define the next phase of it on its own terms. The architecture it describes, open connections, a shared data layer and agents working under one interface, is a plausible direction for the industry. A conversational layer over existing optimisation tools is also something several vendors could claim, which is why the announcement alone does not settle whether agentic retailing is a genuine shift or a repackaging of capabilities airlines already have.
The test will come with the first named airline running Kairos in production, what its agents are permitted to do without human sign-off, and whether third-party agents can connect through MCP in practice. Until PROS provides those details, Kairos is a credible architecture and a pointed piece of positioning, and the evidence that it changes how airlines price and sell is still to come.
