SunExpress gives its operations teams an AI assistant for NOTAMs and manuals
SunExpress has launched AirMind, an assistant built with NTT DATA on Microsoft Azure OpenAI that answers operations staff from NOTAMs, weather, manuals and daily records, citing the source and date. Separately, UiPath AI agents handle crew hotel bookings and disruption records. The airline's CIO says it has built more than 25 generative AI solutions.
The operating picture
Finding the answer in the documents
Hannover Airport SunExpress Boeing 737-8 MAX TC-SOI (DSC00263). Photograph: MarcelX42 / CC BY-SA 4.0 (cropped to fit). Illustrates the airline, not an AI deployment.
Operations staff work from a large and changing body of text: NOTAMs, weather reports, operations manuals and controlled documents. SunExpress, the Turkish Airlines and Lufthansa joint venture based in Antalya (AYT), built AirMind so staff can ask a question in plain language and get an answer that names its source.
Its CIO, Mustafa Egilmezbilek, told AltexSoft in May 2026 that the airline has built more than 25 generative AI solutions and keeps a human in the loop for critical systems. The most specific published examples sit in office workflows run with UiPath.
Selected implementations
The work behind the evidence
01 / Deployed
Answers in plain language, with a source and a date
The operational problem
When a NOTAM, a weather change or a manual revision affects the day's flying, staff need the current rule quickly, and they need to know where it came from.
What the evidence reports
According to NTT DATA's case study, AirMind draws on the airline's Document Control Center, NOTAMs, weather reports, operations manuals and daily operational records, runs on Microsoft Azure OpenAI and answers questions put in plain language within seconds. Each answer carries a source and date reference, and the assistant also raises relevant operational developments without being asked. The case study quotes Murat Erbasan, who leads the operational excellence team in SunExpress's integrated operations control centre.
Operational significance
Citing the source document lets staff check an answer before acting on it, which is the minimum control for a generative model working from regulated material.
Evidence still needed
Neither company has published figures on usage, accuracy or time saved, and the case study gives no launch date. It does not say whether answers are only advisory or feed decisions directly.
Booking hotels for cabin crew and logging the details of disrupted flights are repetitive clerical jobs that build backlogs when the operation is under strain.
What the evidence reports
UiPath's case study says its agents read hotel emails, confirm bookings, update trackers and pass exceptions to staff. They also interpret flight disruptions and send the information to the teams that need it, and group sales uses the system for pricing decisions. Özge Sena Duran, technical product owner at SunExpress, is quoted saying the approach "ensures timely entry of disruption data". UiPath reports $200,000 in savings across the first three processes and a backlog reduced by one to two months, and SunExpress won a UiPath AI25 award in 2025.
Operational significance
These are bounded tasks with a clear route to a human when something does not fit, which is where agent software is easiest to supervise.
Evidence still needed
The savings and backlog figures are UiPath's, with no published baseline or period, and the case study is undated.
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
SunExpress has been open about how it builds AI, including where it keeps a human in the loop. Both documented systems handle information for staff, and neither makes an operational decision on its own.
The more than 25 solutions its CIO describes are not listed publicly. The most useful next disclosure would be AirMind's usage and accuracy in operations control, where an answer drawn from the wrong NOTAM would matter most.
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.
.jpg)