Houcem Fehri
Data Scientist at Aquila Engineering LLC- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Credentials
-
Data Science Foundations - Level 1
IBM
Experience
-
Aquila Engineering LLC
-
United States
-
Oil and Gas
-
1 - 100 Employee
-
Data Scientist
-
Jan 2022 - Present
Working on the Aquila Edge Kiosk: - Improved the response time of the Analytics REST Api from ~10s to 600ms per request using postgresql for storing data, Redis for caching, Asyncio for concurrent processing and FastApi for serving. - Implemented a dashboard to view the analytics using Plotly and Dash. - Created a filtering approach for time series data based on speckle filtering in computer vision, to improve the quality of the sensor signals coming from the rig. - Deployed the project using Docker on the Edge device. Technologies used: Python, Redis, FastApi, Postgresql, Asyncio, Scikit-learn, Numpy, Plotly, Dash
-
-
-
University of Michigan-Dearborn
-
United States
-
Higher Education
-
700 & Above Employee
-
Graduate Research Student
-
Jun 2020 - Apr 2022
Working in BimiLab under the supervision of Dr Jaerock Kwon on Internal Simulation Networks for Autonomous Vehicles :- Collecting data from Carla Simulator and preprocessing it- Using the data to train a neural network system that predicts what the vehicle will see next using the current image and motor information from the vehicle- Applying the system to create a lane keeping system Used SciPy stack, Tensorflow 2.0 and Pytorch in a later iteration.
-
-
Graduate Student Instructor
-
Sep 2019 - Apr 2022
Worked as a Teaching Assistant on multiple graduate and undergraduate courses, my duties include:- Providing office hours for students- Grading the assignments and examinations- Proctoring examinations
-
-
Graduate Research Assistant
-
Sep 2019 - Jun 2020
Worked in ISELab with Dr Marouane Kessentini on the optimization of Software Refactoring in the decision and objective spaces through extracting the developer's preferences:- Implemented a Clustering Algorithm using Expectation Maximization Algorithm for Gaussian Mixture Models- Contributed in writing a paper published in Transaction in Software Engineering
-
-
-
-
Visiting Research Fellow
-
Feb 2019 - Jun 2019
A 4 month fellowship at ISELab in the University of Michigan - Dearborn: - Worked under Dr. Marouane Kessentini to learn about improving the quality of software architecture using A.I - Implemented in Java an algorithm for the transformation from multi objective optimisation to single objective optimisation - Contributed in writing a publication published in SCAM'19 A 4 month fellowship at ISELab in the University of Michigan - Dearborn: - Worked under Dr. Marouane Kessentini to learn about improving the quality of software architecture using A.I - Implemented in Java an algorithm for the transformation from multi objective optimisation to single objective optimisation - Contributed in writing a publication published in SCAM'19
-
-
-
Le Centre d'Etudes et de Recherche des Télécommunications
-
Information Technology & Services
-
Internship Trainee
-
Jun 2018 - Aug 2018
Detecting Suspicious Actions using Deep Learning: - Collected Data using a Raspberry Pi Camera system - Preprocessed Data using OpenCV and Scikit Image - Trained and infered a 3D convolutional network using Keras and Tensorflow Detecting Suspicious Actions using Deep Learning: - Collected Data using a Raspberry Pi Camera system - Preprocessed Data using OpenCV and Scikit Image - Trained and infered a 3D convolutional network using Keras and Tensorflow
-
-
-
Thyna Petroleum Services
-
Oil and Gas
-
1 - 100 Employee
-
Summer Intern
-
Jun 2017 - Jul 2017
-
-
Education
-
University of Michigan-Dearborn
Master's degree, Mechatronics, Robotics, and Automation Engineering -
Tunisia Polytechnic School
National Engineering diploma from l'Ecole Polytechnique de Tunisie, Signals and Systems -
IPEIS - Institut Préparatoire aux Études d'Ingénieur de Sfax
Diplôme d'études universitaires de premier cycle, Physics and Chemistry