Resume
Akrem GOMRI
Data Scientist
13/04/1999
Paris, île-de-France
Data Scientist
13/04/1999
Paris, île-de-France
EDUCATION
-
Paris-East Créteil University
Master in Artificial Intelligence and Cyber-Physical Systems (IA2S)
2023 - 2025 -
Higher Institute of Applied Sciences and Technology
Computer Engineering, Software Engineering
2020 - 2023 - Higher Institute of Applied Sciences and Technology
Preparatory Studies in Math, Physics, Computer Science
2018 - 2020
Paris, France
Sousse, Tunisia
Sousse, Tunisia
EXPERIENCE
- TotalEnergies Digital Factory
- Applied signal analysis and signal processing to time series data from 33 wells internationally, each with 20 sensor variables, recorded over a 3-year period with data taken every 5 seconds.
- Tested different supervised and unsupervised approches for annomaly detection using Deep Learning and Machine Learning approches (Temporal Concolutional Networks, MAT-GAT an Attention graph based algorithm, Tsfresh with Keamns).
Paris, France
PERSONAL PROJECTS
- Hackathon Project: Music Streams Forecast
- Automated the extraction of streaming data for 25,000 musical artists.
- Fine-tuned TimeGPT to forecast future music streams for different profiles.
- Built a Streamlit app to visualize forecasts and evaluate profiles’ potential. Click to test the app
- Predicted which music profiles were worth investing in by estimating their future streaming performance.
- Keywords: TimeGPT (foundation model and a pre-trained transformer for time series), Model Fine-Tuning, Python, Pandas, Numpy, Plotly, Streamlit. Click for more details
- Human Activity Recognition using Video Classification
- Several approaches were used and compared: Simple CNN, CNN + LSTM, pose detection, 3D CNN, Slow Fusion technique.
- Tools: Pytorch for pose detection and TensorFlow for the rest.
- Human Activity Recognition using Wearable Sensors
- Extracted the dataset, then used the sliding window method to capture the pattern with a 2d CNN on the data, then used a layer using different ML approaches : Feed Forward Neural Network, SVM, Decision Tree, etc.
- tools: TensorFlow.
Onsite
Technical Skills and Interests
- Languages: Python, Javascript
- Development Tools: Git, Docker
- Frameworks Libraries: Pytorch, TensorFlow, React, Django
- BigData & Cloud: Spark, DataBricks, Amazon Web Service (AWS)
- Databases: MongoDB, PostgreSQL
- Behavioral Skills: Research-oriented, Competitive, Curious, Autonomous
- Courses Taken: Probability Statistics, Machine Learning, Deep Learning
- Fields of Interest: Natural Language Processing, LLMs, Computer Vision
Certifications
- Professional Certificate in AI Engineering, IBM - MUST University May 2023 - Nov 2023
- French test (TCF): C1 : 559 Sep 2022
- Certificate of Completion in Full Stack Js Course, RedStart-Certificate - Tunis Jun 2021 - Aug 2021
- Certificate of Achievement in Artificial Intelligence, GoMyCode - Tunis Jun 2020 - Aug 2020