Traveologist

For the best experience, view the website on desktop computer.

Chief Traveology
The process

Chief Traveology Officer

The first C-level executive completely made out of data

Read the position Reveal the officer
Read the position
watch ()

Interview with Pascal Clement, Head of Travel Intelligence

reveal

The first Chief of Traveology

The Position

Everyday humans generate petabytes of data. Data is in a constant evolution, just like humans. This is why we created and hired the first and only data-human who has a unique perspective of travel.

Position Summary

Assigned as the first in history Chief Traveology Officer, 
Shin Walker will deliver a socio-anthropological view on humankind as a traveling species.

Representing the globality of the travel data, he will be able to improve any decision based on the fact that he has all the data of all the travelers. He’s bringing the power of data to every process around travel.

His role will be to work closely with all Amadeus functions to enhance processes internally and within our solutions in order to improve the traveler experience, as well as our customers efficiency and bottom line.

Reveal the first Chief of Traveology
watch ()

Interview with Pascal Clement, Head of Travel Intelligence

reveal

The first Chief of Traveology

@ 3 | # Y & O J { 6 ( M 5 D 2 E V / 7 R

The Process

The Data

Shin gives Amadeus a view of how raw data can be transformed into actionable insights and through conducting the analysis of 3.4 billion rows of flight data which were used to create Shin, interesting traveler trends have been revealed.

The Amadeus Airlines Data Unit collaborated with the Research and Innovation team to firstly, extract aggregated demographic information using big data technologies, and then generate a face to match this data using machine learning and artificial intelligence.

We collected booking data from all the different airlines and travel agencies available in our systems and then merged it.

We used anonymization, aggregation and hashing to make it impossible to track any individual passenger.

We then computed the desired statistics, finding the average distribution of age, gender and nationality / ethnicity.

Two advanced machine learning algorithms were then used in tandem to create a face form an open database of 10,000 adult face images.

One of them was creating thousands of fake images based on face tagged database (the generator) and another one was identifying whether the produced images were real or not (the discriminator).

By letting the two algorithms play and communicate with each other, we reached an equilibrium state where the discriminator learned to recognize real human faces and by doing so, helped the generator to create realistic content (Shin’s face).