Makia Zmitri

Customer Success Data Scientist at Amiral Technologies
  • Claim this Profile
Contact Information
us****@****om
(386) 825-5501
Location
Grenoble, Auvergne-Rhône-Alpes, France, FR
Languages
  • Français Native or bilingual proficiency
  • Anglais Full professional proficiency
  • Espagnol Elementary proficiency
  • Arabic Native or bilingual proficiency
Skills

Topline Score

Topline score feature will be out soon.

Bio

Generated by
Topline AI

You need to have a working account to view this content.
Join now
You need to have a working account to view this content.
Join now

Credentials

  • Qualification for Assistant Professor positions
    Section 61 - Génie informatique, automatique et traitement du signal
    Feb, 2022
    - Nov, 2024
  • Neural Networks and Deep Learning
    DeepLearning.AI
    Aug, 2020
    - Nov, 2024

Experience

  • Amiral Technologies
    • France
    • IT Services and IT Consulting
    • 1 - 100 Employee
    • Customer Success Data Scientist
      • Apr 2023 - Present

      - Conduct feasibility studies on customer use cases from start to finish - Train customers on machine learning concepts - Coach customers on how to use the company's software and create predictive models - Work closely with the development team to integrate customer needs collected in the field

    • Data Scientist
      • Oct 2021 - Apr 2023

      - Develop DiagFit, a software allowing failure prediction of industrial equipment. - Design innovative algorithmic bricks resulting from various research works. - Integrate new modules and improve the quality of internal code. - Take responsibility of feasibility studies in pilot projects.

  • GIPSA-lab
    • France
    • Research Services
    • 1 - 100 Employee
    • PHD Candidate
      • Oct 2018 - Sep 2021

      Title: Magnetometer Array-Based Indoor Navigation Using Kalman Filter Abstract: This thesis focuses on GPS-free indoor navigation using only strapdown low-cost magneto-inertial sensors and without relying on any prior-mapping of the magnetic field, nor on a dedicated infrastructure. The main idea is to take advantage of the magnetic field’s disturbances present indoor to generate not only a velocity estimate, but also attitude, and position of the moving body under study, enabling therefore… Show more Title: Magnetometer Array-Based Indoor Navigation Using Kalman Filter Abstract: This thesis focuses on GPS-free indoor navigation using only strapdown low-cost magneto-inertial sensors and without relying on any prior-mapping of the magnetic field, nor on a dedicated infrastructure. The main idea is to take advantage of the magnetic field’s disturbances present indoor to generate not only a velocity estimate, but also attitude, and position of the moving body under study, enabling therefore the reconstruction of any performed trajectory. To do so, the spatial derivatives of the magnetic field are explored throughout a specific set of a magnetometer array. From this array, the magnetic field gradient is determined using approximation methods. It follows that it suffers from uncertainties and noise. For this reason, a standard Magneto-Inertial Navigation (MINAV) model is enhanced by introducing a new equa- tion that describes the magnetic field gradient dynamics. The new proposed model stands out from the usual ones used in the corresponding literature, as it fully cap- tures the richness of the magnetic field gradient variations, and enables reducing its uncertainties and noise. Then, different algorithms based on Extended Kalman Filtering (EKF) are implemented, to make use of the proposed model. Nevertheless, the performance of the EKF is degraded under certain conditions, mostly related to measurements quality. Therefore, it becomes necessary to combine it with the Zero-velocity Update Technique (ZUPT), in the case of foot-mounted navigation or Deep Neural Networks (DNNs) in the more general case. The proposed algorithms are assessed not only on simulated data but also on a real experimental benchmark using a sensor array, in presence of ground truth equipment. The obtained results illustrate the contribution of this thesis on the velocity estimation and consequently on the trajectory reconstruction. Show less Title: Magnetometer Array-Based Indoor Navigation Using Kalman Filter Abstract: This thesis focuses on GPS-free indoor navigation using only strapdown low-cost magneto-inertial sensors and without relying on any prior-mapping of the magnetic field, nor on a dedicated infrastructure. The main idea is to take advantage of the magnetic field’s disturbances present indoor to generate not only a velocity estimate, but also attitude, and position of the moving body under study, enabling therefore… Show more Title: Magnetometer Array-Based Indoor Navigation Using Kalman Filter Abstract: This thesis focuses on GPS-free indoor navigation using only strapdown low-cost magneto-inertial sensors and without relying on any prior-mapping of the magnetic field, nor on a dedicated infrastructure. The main idea is to take advantage of the magnetic field’s disturbances present indoor to generate not only a velocity estimate, but also attitude, and position of the moving body under study, enabling therefore the reconstruction of any performed trajectory. To do so, the spatial derivatives of the magnetic field are explored throughout a specific set of a magnetometer array. From this array, the magnetic field gradient is determined using approximation methods. It follows that it suffers from uncertainties and noise. For this reason, a standard Magneto-Inertial Navigation (MINAV) model is enhanced by introducing a new equa- tion that describes the magnetic field gradient dynamics. The new proposed model stands out from the usual ones used in the corresponding literature, as it fully cap- tures the richness of the magnetic field gradient variations, and enables reducing its uncertainties and noise. Then, different algorithms based on Extended Kalman Filtering (EKF) are implemented, to make use of the proposed model. Nevertheless, the performance of the EKF is degraded under certain conditions, mostly related to measurements quality. Therefore, it becomes necessary to combine it with the Zero-velocity Update Technique (ZUPT), in the case of foot-mounted navigation or Deep Neural Networks (DNNs) in the more general case. The proposed algorithms are assessed not only on simulated data but also on a real experimental benchmark using a sensor array, in presence of ground truth equipment. The obtained results illustrate the contribution of this thesis on the velocity estimation and consequently on the trajectory reconstruction. Show less

  • Université Grenoble Alpes
    • France
    • Higher Education
    • 700 & Above Employee
    • University Teacher
      • Jan 2020 - Jun 2021

      - DUT Génie Électrique et Informatique Industrielle (GEII) - IUT Grenoble 40 hours of Mathematics for "License 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 15 hours of SISO Feedback Control for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 3 hours of state representation for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 12 hours of control theory… Show more - DUT Génie Électrique et Informatique Industrielle (GEII) - IUT Grenoble 40 hours of Mathematics for "License 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 15 hours of SISO Feedback Control for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 3 hours of state representation for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 12 hours of control theory for "License 3" students Show less - DUT Génie Électrique et Informatique Industrielle (GEII) - IUT Grenoble 40 hours of Mathematics for "License 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 15 hours of SISO Feedback Control for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 3 hours of state representation for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 12 hours of control theory… Show more - DUT Génie Électrique et Informatique Industrielle (GEII) - IUT Grenoble 40 hours of Mathematics for "License 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 15 hours of SISO Feedback Control for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 3 hours of state representation for "Master 1" students - UFR Physique Ingénierie Terre Environnement Mécanique (PhITEM) – UGA 12 hours of control theory for "License 3" students Show less

  • GIPSA-lab
    • France
    • Research Services
    • 1 - 100 Employee
    • Final Engineering Year/Master Internship
      • Mar 2018 - Jul 2018

      - State estimation for the determination of attitude using inertial measures fusion - Application to the capture and analysis of human movements - Classification and prediction with machine learning algorithms - State estimation for the determination of attitude using inertial measures fusion - Application to the capture and analysis of human movements - Classification and prediction with machine learning algorithms

  • Power Sun Energy - PSE
    • Tunisia
    • Renewables & Environment
    • 1 - 100 Employee
    • Engineer Internship
      • Jun 2017 - Sep 2017

      - Critical analysis of the dimensioning of a photovoltaic storage system and the development of an optimization solution. - Research on the requirements and certifications requested by the National Agency of Energy Control (ANME) and the Tunisian Company of Electricity and Gas (STEG). - Development of a suppliers database of different solar products. - Organizing meetings with international partners and presenting projects. - Critical analysis of the dimensioning of a photovoltaic storage system and the development of an optimization solution. - Research on the requirements and certifications requested by the National Agency of Energy Control (ANME) and the Tunisian Company of Electricity and Gas (STEG). - Development of a suppliers database of different solar products. - Organizing meetings with international partners and presenting projects.

  • Shams Energy Access
    • Tunisia
    • Renewables & Environment
    • 1 - 100 Employee
    • Worker Internship
      • Jul 2016 - Jul 2016

      - Identifying customer requirements and preparing the PV design using the SAP ERP software system. - Going on site for solar photovoltaic system installations. - Verifying the functionality of the system through real time testing. - After-sales services, maintenance and troubleshooting. - Advertisement and marketing activities. - Storage and inventory management. - Identifying customer requirements and preparing the PV design using the SAP ERP software system. - Going on site for solar photovoltaic system installations. - Verifying the functionality of the system through real time testing. - After-sales services, maintenance and troubleshooting. - Advertisement and marketing activities. - Storage and inventory management.

Education

  • Université Grenoble Alpes
    PhD, Automatics and Production
    2018 - 2021
  • National Engineering School of Tunis (ENIT), University of Tunis El Manar
    Master's degree (M2), Automatic Control
    2017 - 2018
  • National Engineering School of Tunis (ENIT), University of Tunis El Manar
    Engineer's degree, Electrical Engineering, Automatics
    2015 - 2018
  • Preparatory Institute for Engineering Studies - El-Manar (IPEIEM)
    A degree in university studies of the first two years, Mathematics and Physics
    2013 - 2015

Community

You need to have a working account to view this content. Click here to join now