Event

Pitch Day – Pi School of AI Session 10

Pitch Day Pi School of AI Session 10

Watch the School of AI fellows present the process, hurdles and results of their challenges.

  • After working for 8 weeks, the fellows have made incredible progress in solving the challenges posed by Octo Telematics, Translated, Neural Jam and all the other sponsors. We are grateful to these partners for their support and commitment to innovative solutions. Are you curious to find out how the fellows did? Join us!
Register here

Our Pi School of AI Fellows will pitch these challenges:

Challenge 1 - Predicting balance sheet performance for private equity investment targeting

  • Challenge 1

    Predicting balance sheet performance for private equity investment targeting
    Sponsor: Ardian Growth
    This challenge focused on the complex problem of business forecasting. More specifically, it focused on predicting whether a company will be a good investment or acquisition target. The Pi School fellows worked hard to create a system that could use historical financial data from a company to predict how it would be performing in the future!

  • Deressa Wodajo

    Fellow

  • Eric Walzthöny

    Fellow

Challenge 2 - AI for accurate tank level estimation in rental cars

  • Challenge 2

    AI for accurate tank level estimation in rental cars
    Sponsor: Octo Telematics
    An important part of missing profit for car rental companies is inaccuracies in fuel gauge readings. As a concrete example, a customer might pick up a car with a full tank from the car rental company and drive for 20 kilometers before returning it. If the fuel gauge still reads 100% when the vehicle is returned, the customer cannot be charged any amount even though they drove for 20 kilometers. In this challenge, we use AI to accurately estimate the tank level based on a history of OBD (On-Board Diagnostics) sensor readings spanning the time frame between when the customer picks up the car and when they return it.

  • Lloyd Thompson

    Fellow

  • Sinan Gencoglu

    Fellow

  • Burhan Hussein

    Fellow

  • Hue Salari

    Fellow

Challenge 3 - Measuring driving distance without distance sensors

  • Challenge 3

    Measuring driving distance without distance sensors
    Octo Telematics
    Octo Telematics is interested in doing risk scoring of drivers' behaviour, and driving distance is an important factor in determining such a score. Interestingly enough, virtually all modern Smartphones are equipped with GPS and inertial sensors (accelerometer, gyroscope, magnetometer), which make them ideal for this task. In this challenge, however, we go one step further. Given that using GPS poses several concerns to the end-user (battery life and privacy, to name a few), we set out to use only inertial sensors to estimate the driving distance.

  • Muhammad Ghauri

    Fellow

  • Nirzaree Vadgama

    Fellow

Challenge 4 - AI augmented medical seminars

  • Challenge 4

    AI augmented medical seminars
    Conferences and events in the medical sector often involve numerous speakers, presentations and scientific works. In this challenge, we use Natural Language Processing to make sense of the sizable amount of material that is the output of such conferences by drawing the most relevant insights for both participants and sponsors.

  • Stefano Fiorucci

    Fellow

  • Jaume Francesca

    Fellow

Challenge 5 - AI personality profiling for professional development.

  • Challenge 5

    AI personality profiling for professional development
    Sponsor: Neural Jam
    Neural Jam Intelligent Pathfinder (IPF) is a profiling tool to synthesize values, work styles, and interests for any individual. Our work leverages AI to connect IPF profiles with the landscape of professions. We built a tool to identify the group of professions closer to your values and interests, providing a list of skills important to master in order to build your personal development plan.

  • Serdar Altan

    Fellow

  • Luigi Palumbo

    Fellow

Challenge 6 - Call center analytics

  • Challenge 6

    Call center analytics
    This challenge consists of two complex tasks. Task 1 concerns the elaboration of guidelines that some transcribers will use to convert audio files into text. The resulting dataset will be used to improve an automatic speech recognition (ASR) tool. To evaluate this tool, the team developed suited metrics that will help to understand the quality of automatic transcriptions compared to human ones. On the other hand, task 2 requires the extraction of the reason of call (ROC) of some transcribed calls between operators and customers. The team adopted state-of-the-art technologies, customizing innovative models of Natural Language Processing to be tailored to the technical and peculiar language used in these phone calls.

  • Marcello Politi

    Fellow

  • Irene Migliore

    Fellow

  • Michela Pascale

    Fellow

Pi School of AI Mentors

  • Federico Wolenski

    Machine Learning Engineer for NLP applications in Almawave S.p.A.

    Federico Wolenski is a Mathematician by training with several years of experience working as a Machine Learning Engineer for NLP applications in Almawave S.p.A..
    He earned his Ph.D. working on research topics in the broad field of Geometry and now applies his knowledge to the research and development of advanced Deep Learning Models.

  • David Preti

    ML Research Engineer in Almawave S.p.A

    David Preti holds a Ph.D. in Theoretical Particle Physics. He has six years of experience on numerical and theoretical aspects of lattice quantum field theories.
    David worked as a Ph.D. candidate at the Instituto de Fisica Teorica (CSIC/UAM) in Madrid and later as a postdoctoral researcher at Istituto Nazionale di Fisica Nucleare (INFN) in Turin.
    Along with physics, David specialized in Deep Learning. He currently works as an ML Research Engineer in Almawave S.p.A., where his responsibilities range from applied research to development and implementation of AI approaches in the context of Natural Language Processing.
    In addition to professional interests, David has a passion for domotics, martial arts, water sports and swing dance.

  • Fabio Fumarola

    Principal AI Engineer Computer Vision and Machine Learning at Prometeia

    Fabio Fumarola got a Ph.D. in Machine Learning (ML) in 2011, and he has been working in this field for several years. He applied ML techniques in different areas ranging from natural language processing, graph mining, computer vision, stream analysis and representation learning to problems related to finance, gaming and the public sector.

Register for the Pitch Day event

We are sorry to inform you that registration for the in-person event is closed. Please, send us an email to join the waiting list. But, you can still register for the online event here! We will send you the event link via email.

  • Hidden

Do you have a challenge that AI can solve?

  • AI for Business Register for the Pitch Day event to understand what Pi School can do for you.
REGISTER NOW