SriLankan credits AI revenue management in a 13% revenue rise
SriLankan Airlines says a new revenue management system using AI and machine learning was among the drivers of 13% revenue growth in the first three quarters of its 2025-26 financial year. Its GPT-4 chatbot, Yaana, has answered passengers on the airline's website since February 2025. Both sets of figures are the airline's own.
The operating picture
Pricing the whole journey
158ai - SriLankan Airlines Airbus A340-311; 4R-ADB@LHR;27.10.2001. Photograph: Aero Icarus / CC BY-SA 2.0 (cropped to fit). Illustrates the airline, not an AI deployment.
A conventional revenue management system controls availability flight by flight. An origin and destination system values a seat by the passenger's whole journey, which matters for a carrier selling connections through its Colombo (CMB) hub. In March 2026 SriLankan said it had moved to PROS Revenue Management Advantage with Amadeus Revenue Availability and Active Valuation, which it describes as AI and machine learning powered.
Its AI for passengers arrived a year earlier. Yaana, built with CodeGen International, answers passenger questions on the airline's website using GPT-4, drawing on SriLankan's policy documents and live schedule data.
Selected implementations
The work behind the evidence
01 / Deployed
Valuing a seat by the passenger's whole trip
The operational problem
A seat on a flight into Colombo can be worth more sold to a passenger connecting onward than to someone ending their journey there. Controls set flight by flight cannot see that difference.
What the evidence reports
SriLankan said on 27 March 2026 that the PROS and Amadeus platform supports dynamic pricing, inventory optimisation, real-time seat availability and predictive analytics across its sales channels, including premium seat allocation. It reported 13% revenue growth in the first three quarters of the 2025-26 financial year and called the system "a key driver, among other strategic factors". Kshanaka Saparamadu, head of revenue management, said the move refined the airline's existing processes.
Operational significance
Few airline AI announcements are tied to a financial result. The airline's own wording also limits this one: the system is one driver among several, and SriLankan does not say how large its share was.
Evidence still needed
The release gives no start date, no comparison period for the 13% and no split between pricing and other causes such as capacity or demand. It cites an industry benchmark of 3% to 5% revenue gains from origin and destination systems, which is not a SriLankan measurement.
Passengers contact the airline at all hours about bookings, policies and disruption, and during irregular operations they need rebooking options quickly.
What the evidence reports
SriLankan said on 6 February 2025 that Yaana, developed with CodeGen International, uses a GPT-4 virtual assistant with retrieval from policy documents and real-time data such as flight schedules, and helps passengers find alternatives during disruption. Dimuthu Tennakoon, head of worldwide sales and distribution, said it had handled close to 12,000 inquiries since going live, "autonomously resolving 88% of them".
Operational significance
Answers drawn from the airline's own policy documents and schedules are the main protection against a chatbot stating a rule that does not exist.
Evidence still needed
The release does not give the period behind the 12,000 inquiries, define "resolved" or list the languages supported. No later usage figures have been published.
A source can verify an announcement without verifying its projected outcome.
The wider register
Other tracked activity
Other records are shown separately, including non-AI technology and programmes at different stages of deployment. A register entry does not establish a measured AI benefit.
No other publicly listed records.
Editorial assessment
SriLankan's two systems sit at either end of the commercial chain: one sets the fare, the other answers the passenger who bought it. Both are described in the airline's own releases, with named executives and figures.
Those figures need context to carry weight. The revenue claim needs a comparison period and some measure of the system's share; the chatbot claim needs a definition of what Yaana counts as resolved.
Live data from the Aerostack register. Status reflects published evidence only. See Our approach for how records are graded.
Aerostack register · 2 tracked deployments
Aerostack tracks AI deployments across airports and airlines, with the published evidence behind each record. Photography is real aviation imagery. Record thumbnails may show the named airport or airline; other photographs do not.
