MENTHOR Ai is pilot training management software that checks every rostered crew member, cabin crew included, against the qualifications their flight requires, turns recurring FDM events into training and records EBT competency assessments and line training. Check-in for the first flight of the day stays closed until that flight's whole crew is compliant. Keyword rules match approved FDM events to a course, and the training team sees each pilot only as a numbered crew member.
Email me the brochure, MENTHOR Ai
MENTHOR Ai is FlightAtom's pilot training management software for training managers, instructors, safety teams and pilots. Before each flight it checks every rostered crew member, cabin crew included, against the qualifications that flight requires, and keeps check-in for the first flight of the day closed until that flight's whole crew is compliant. Its Auto Training matches recurring, reviewer-approved FDM events to a course with keyword rules and shows the training team each pilot only as a numbered crew member; the course is assigned by the training team or, once an administrator switches automation on, automatically, and MENTHOR Ai then compares the count of the triggering event type before and since. MENTHOR Ai also records EBT competency assessments and line training, and delivers courses that need a pass of at least 80%.
Tomorrow's flights are rostered, and Safety has approved the latest FDM events. The training team needs to know who is qualified to fly and who needs training, without singling anyone out.
MENTHOR Ai checks each crew member's required qualifications against the flights they are rostered on, flags airport, visa and right-hand-seat endorsements, and shows today's and tomorrow's readiness on one board.
Auto Training finds the event types that recur for a pilot in the approved FDM events and matches a course with keyword rules, showing the pilot only as a numbered crew member.
The training team assigns the course, or MENTHOR Ai assigns it once automation is on. The pilot is told it is developmental, not disciplinary, and the efficacy view compares the events before and since.
A readiness check for every rostered flight, and a course assignment that records the triggering event type, who assigned it and when, with completion and the before-and-after count kept alongside.
Safety approves the events. A training administrator assigns or dismisses each finding, or switches on automatic assignment, and names appear only in the authorized safety view.
MENTHOR Ai turns recurring, approved FDM events into individual training. Safety and Training open each finding through separate access checks, so the training team can assign a course without seeing who the pilot is.
Explore the current product interface with demonstration data. Open any screenshot for a closer look.
The Daily Crew Readiness board lists today's flights and gives every crew member a readiness light and a check-in mark: green when nothing their rank requires is missing or expired, blue for a new captain who may fly only with an instructor, and red for a first officer whose OPC is out of date. US visa and right-hand-seat chips appear where a flight needs them. Synthetic demonstration data.
On the captain's flight page, the rest of the crew each carry a readiness light and a US visa chip, so a first officer who is not mission-ready shows red beside the route, fuel plan and airport weather. Synthetic demonstration data.
On the captain's phone, check-in for the next flight stays closed because the first officer is not compliant, and the next-flight panel shows the rest of the crew's readiness and US visa status. Synthetic demonstration data.
Auto Training's fleet statistics for the last 180 days show crew-attributed FDM events by type and by seat, how many pilots have how many events, and a ranking that names pilots only as numbered crew members. Synthetic demonstration data.
Each row is a recurring FDM event type for one de-identified pilot, with its count, severity and the course a keyword rule recommends. In manual mode the training team assigns or dismisses each finding. Synthetic demonstration data.
For every FDM-triggered course, the count of the same event type in the 180 days before the assignment and since, with course completion; rows show fewer, unchanged and more events since. Crew numbers are assigned per view, so the same pilot can carry a different number on another screen. Synthetic demonstration data.
Fleet FDM indicators normalised per 1,000 operations against alert levels derived from the operator's own data, with an 18-month control chart that marks out-of-control months. Synthetic demonstration data.
A first officer sees their own approved FDM events per 1,000 operations over 180 days against the first-officer average, a trailing three-month trend, repeated event types and each event with a button to add their own account, once Safety opens the view to them. Synthetic demonstration data.
A signed OPC/LPC competency assessment: each competency is graded 1 to 5 or marked not observed, the grade of 2 carries a written observation, and the record shows the average and the trainee's acknowledgement. Synthetic demonstration data.
MENTHOR Ai works from the operator's own training records, flight schedule and approved FDM events. Auto Training connects the safety and training workspaces without showing the training team who the pilot is, and pilots have their own app, on the web and on iPhone and iPad, for their qualifications, flights with crew readiness, briefings, logbook, check-in and courses.
MENTHOR Ai checks each crew member's required qualifications by rank, cabin and senior cabin crew included, against the flights they are rostered on. It flags the airport endorsements a destination or its alternate needs, a visa where the destination requires one and a right-hand-seat qualification for a captain rostered in the first officer's seat, shows new pilots still completing line training as able to fly with an instructor, and puts today's and tomorrow's readiness on one board for training, flight operations, operations control and maintenance control.
When check-in opens, two hours before a crew member's first flight of the day, MENTHOR Ai checks everyone on that flight against the same readiness rules and the flight's airport, visa and right-hand-seat requirements. If anyone is missing an item or out of date, check-in stays disabled on the web and in the iPhone and iPad app, with a message naming who and why, and check-in reminders stop. Each crew member's dashboard also shows the readiness of the rest of the crew on their next flight.
Auto Training reads the FDM events Safety has approved, finds the event types that recur for each captain or first officer, and routes the finding to the matching course in the MENTHOR Ai course library. A fleet statistics view shows crew-attributed events by type, severity and seat, how they are spread across pilots and how many findings are ready to assign. Each finding shows its event count, last occurrence and average and maximum severity, and a dismissed finding stays dismissed.
The training view identifies pilots only as Crew member 1, Crew member 2 and so on, with no name, rank or seat position, and names and crew IDs are never sent to it. A training administrator assigns a course through an encrypted reference that only the server can resolve. Authorised safety staff open the same queue with names beside the labels through a separate endpoint and access check, and Auto Training results stay de-identified in the LMS views and issue no certificate.
Each Auto Training course has a seven-day completion window, a daily countdown reminder and one escalation to the training team and chief pilot if it goes overdue, which never names the pilot. For each assignment, MENTHOR Ai then shows how many events of the same type the pilot had before and since, with a trend marker and the completion status. The pilot is notified in the app and by email, and the email explains the rules-based process behind Auto Training and that the course is developmental, not disciplinary.
MENTHOR Ai starts in manual mode: a training administrator reviews each finding with the suggested course pre-selected and assigns it, assigns a different course or dismisses it. Switching automation on means confirming roughly how many open findings will be assigned straight away; MENTHOR Ai then assigns that backlog and picks up newly qualifying findings every five minutes. The same course is never assigned to a pilot twice automatically, and each assignment records who made it, when, and the triggering event type and reason.
The pairing planner proposes the OPC and LPC/OPC simulator plan that pairs the most pilots inside their 90-day windows and each month's simulator capacity, schedules overdue checks at the earliest opportunity and sends approved sessions to a shared training calendar. Upcoming expiries are grouped into training cohorts with an estimate of the training days needed. Courses open page by page, with at least 30 seconds on every page before the assessment unlocks, count as complete only with a pass of at least 80%, and issue PDF certificates; authority staff can have accounts limited to their assigned courses.
Instructors grade the ICAO pilot competencies, plus pilot monitoring, from 1 to 5 or as not observed, for sessions such as OPC, LPC, CRM and ground school. A grade of 2 or below needs a written observation, the server recomputes the average grade, and the trainee countersigns a record that cannot be edited. Each development point is checked against the next assessment, and every trainee has one cumulative line-training record that each line instructor adds to, with a sector log, item grades and progress toward the minimum sector count.
Safety uploads the FDM provider's event export, and MENTHOR Ai matches each validated event to the operated flight by aircraft, time and station, attributes it to the flight-deck crew and queues ambiguous matches for review before Safety approves the batch. Safety can remove an event from a pilot's record or reassign it to another crew member, with an audit entry, and each pilot sees only their own approved events, can respond to each, and on a Trends page follows their own events per 1,000 operations over a rolling 180 days against the average for their seat. Fleet indicators are tracked per 1,000 operations against alert levels derived from the operator's own data, with CUSUM and EWMA control charts to flag sustained change, a projection of when an alert level could be reached, and an analysis view that compares the last 90 days with the prior 90 and breaks events down by phase, station and fleet.
MENTHOR Ai's checks and course assignments are deterministic: keyword rules route recurring FDM events to courses, statistical tests flag safety drift, and an optimisation algorithm proposes simulator pairings. DELPHI Ai writes the monthly crew coaching, and external AI assistants connect through scoped, revocable credentials.
Auto Training uses only FDM events that a reviewer has approved, that are not on an excluded list and that are attributed to a captain or first officer, counting each event once per pilot. A finding is raised when the same event type occurs at least twice within the default 180-day window, or on the first occurrence for an unstable approach. Ordered keyword rules then pick the course, and the first match decides: TCAS events go to TCAS resolution, GPWS and terrain alerts to EGPWS, and unstable approaches, high sink rates and hard landings to stabilized approach, with further rules for overspeeds, long flares, taxi speed, stall, upset recovery, windshear and ATC phraseology. An event type that matches no rule gets no recommendation.
The planner compares several pairing strategies, including a maximum-matching algorithm, and keeps the plan that pairs the most pilots within their 90-day windows and each month's simulator capacity. LPC/OPC sessions keep matching pairing codes, the planner prefers to pair a captain with a first officer, and overdue checks take the earliest opportunity.
On the first of each month, DELPHI Ai writes two to four coaching recommendations for each active pilot from a de-identified profile of their approved FDM events, and for cabin crew from their passenger-survey ratings, citing the operator's manuals. Names never leave MENTHOR Ai, and each crew member sees their own coaching. DELPHI Ai also chooses the Pilot of the Month and Cabin Crew of the Month from a shortlist MENTHOR Ai scores, seeing labels instead of names. AI assistants connect through a Model Context Protocol endpoint with scoped, revocable and rate-limited credentials, and the operator can open the DELPHI Ai knowledge assistant to all pilots or to named pilots, who acknowledge the usage policy on each visit, with their questions kept in an audit trail.
Agree a representative sample and a baseline from your current process. These are proposed evaluation measures, not published performance results.
A training requirement, a pilot record, an instructor assessment, and a representative FDM event export.
Follow currency and an Auto Training finding from the recurring event through the de-identified assignment, completion, and the before-and-after count.
Record preparation time; time from a recurring event to an assigned course; overdue assignments; assessment completeness.
Through Auto Training. MENTHOR Ai works from the validated event export of the operator's existing FDM provider and uses only events that a reviewer has approved and that are attributed to a captain or first officer. When the same event type occurs for a pilot at least twice within the default 180-day window, or once for an unstable approach, ordered keyword rules pick the matching course, such as TCAS resolution, EGPWS or stabilized approach. The training team assigns it, or MENTHOR Ai assigns it automatically once an administrator switches that on.
No. The training view of Auto Training identifies pilots only as Crew member 1, Crew member 2 and so on, with no name, rank or seat position, and names and crew IDs are never sent to it. A training administrator assigns a course through an encrypted reference that only the server can resolve. Authorised safety staff see the same queue with names through a separate endpoint and access check, and Auto Training results stay de-identified in the LMS views.
Only when an administrator switches it on. MENTHOR Ai starts in manual mode, where a training administrator reviews each finding with the suggested course pre-selected and assigns it, picks a different course or dismisses it. Switching automation on means confirming roughly how many open findings will be assigned and emailed straight away. MENTHOR Ai then assigns that backlog and picks up newly qualifying findings every five minutes until it is switched off, never assigning the same course to a pilot twice.
For each FDM-triggered assignment, MENTHOR Ai lists the number of events of the same type the pilot had before the assignment and since, with a trend marker and the course completion status. A course counts as complete only after the pilot has studied every page for a minimum time, finished every lesson and passed the assessment with at least 80%. The pilot gets a daily reminder during the seven-day completion window, and an overdue course is escalated once without naming the pilot.
Yes. MENTHOR Ai checks each crew member's required qualifications from the operator's training records, by rank and including cabin crew, against the flights they are rostered on. It adds the airport endorsements a destination or its alternate needs, a visa where the destination requires one and a right-hand-seat qualification for a captain in the first officer's seat. A shared board shows today's and tomorrow's readiness, and check-in for the first flight of the day stays closed until that flight's whole crew is compliant.
Yes. Instructors grade the ICAO pilot competencies, plus pilot monitoring, from 1 to 5 or as not observed, and a grade of 2 or below needs a written observation. The server recomputes the average grade, the signed record cannot be edited, and the trainee countersigns it. The record complements the operator's approved check forms.
Its checks and course assignments do not: Auto Training and the other built-in automation are deterministic. Keyword rules route recurring FDM events to courses, statistical control charts flag safety drift, and an optimisation algorithm proposes simulator pairings. DELPHI Ai writes the monthly coaching, from a de-identified profile of each active pilot's FDM events and of cabin crew passenger-survey ratings, and chooses the Pilot of the Month and Cabin Crew of the Month from a shortlist MENTHOR Ai scores. AI assistants can connect through a Model Context Protocol endpoint with scoped, revocable credentials, and the operator can enable the DELPHI Ai knowledge assistant for pilots, with a policy acknowledgment and an audit trail.
No. MENTHOR Ai has its own pilot app and administration workspaces for training, safety and flight operations, and can be evaluated as a standalone training product. Each deployment connects to the operator's training records, flight schedule and FDM event export.
As your fleet, routes, rules or teams change, we adapt MENTHOR Ai to match, and you approve each change before it goes live.
A focused walkthrough, with the people and decisions that matter to your team.