Houssem Ben Braiek

Houssem Ben Braiek

Ph.D., M.Sc., Eng.

Software Engineering

About Me

I am ML Tech Lead at Sycodal, a company developing visual inspection systems and collaborative robots for small to medium-sized industries. In this role, I supervise and guide the development of machine learning solutions designed to empower software modules with artificial intelligence. I hold a M.sc. and Ph.D. in software engineering from Polytechnique Montreal, where both of my dissertations were awarded the prize for best thesis in computer science. This involves being a resident ML researcher at Bombardier for two years, working on model testing and out-of-distribution detection approaches for tackling reliability and certification challenges of ML models in aircraft engineering problems. I published research papers on assurance quality of machine learning systems published in top-tier scientific journals and prestigious conferences. I participated in designing and developing open-source debugging and testing tools for deep learning programs. I am passionate about staying up-to-date on the latest MLOps initiatives, and I occasionally share my thoughts in technical workshops and blog posts.

Interests
  • MLOps
  • Software Engineering
  • Machine Learning
  • Computer Vision
  • Robotics
Education
  • Ph.D. in Software Engineering, CS Best Thesis Award, 2022

    Polytechnic Montreal

  • M.Sc. in Software Engineering, CS Best Thesis Award, 2019

    Polytechnic Montreal

  • Software Engineer, 1st GPA Rank, 2017

    National Institute of Applied Science and Technology

Experience

 
 
 
 
 
ML Tech Lead
Jan 2023 – Present Montreal
 
 
 
 
 
Formateur
Sep 2023 – Present Montreal
 
 
 
 
 
MLOps Specialist
Sep 2022 – Dec 2023 Montreal
Designing and maturing the ML engineering and deployment workflows for the vision recognition models, data-driven control modules, and predictive maintenance systems integrated into Sycodal’s robot fleet and IoT platform.
Technologies: DVC, MLflow, NVIDIA AI Tools Suite, Databricks, and Azure IoT.
 
 
 
 
 
Applied ML Researcher
Jan 2020 – Aug 2022 Montreal
Building and validating trustworthy ML-powered system simulations to accelerate aircraft development and certification. Technologies: Tensorflow, Pytorch, Scikit-learn, and AzureML.
 
 
 
 
 
SE Researcher
Feb 2019 – Jul 2019 Montreal
Statistical Analysis of uplifted commits to identify what characterizes the risky uplifts that will likely introduce software regressions. Technologies: Pandas, Scikit-learn, Seaborn, and Parsimonious.
 
 
 
 
 
Research Engineer
Apr 2019 – Dec 2019 Montreal
Development of an AI-enabled web application for managing candidates and job offers. Using natural language processing, the integrated AI features retrieve, process, and match textual data from resumes and offer letters in PDF and Word formats. Technologies: Bootstrap, Django, spaCy, and NLTK.
 
 
 
 
 
Research Engineer
Mar 2017 – Aug 2017 Paris
Scaling up of machine learning systems using Big Data processing to deal with high-dimensional, large datasets like remote sensing hyperspectral images; Acceleration of intensive tensor-based computations using NVIDIA GPUs and CUDA libraries. Technologies: Hadoop, Spark, Tensorflow, and CUDA.
 
 
 
 
 
Business Intelligence Developer
Jun 2016 – Sep 2016 Tunis
Designing and Implementing a cloud-hosted BI solution for account management activity reports (CRA).
Technologies: SQL Server, MSBI Suite, and Power BI.
 
 
 
 
 
Enterprise Solution Developer
Jun 2015 – Sep 2015 Tunis
Designing and implementing a football team’s FRM solution (Fan Relationship Management).
Technologies: MS Dynamics CRM (BackOffice) and MS Sharepoint (FrontOffice).

Publications

(2024). In-Simulation Testing of Deep Learning Vision Models in Autonomous Robotic Manipulators. In ASE'24.

Cite Arxiv

(2024). Machine Learning Robustness: A Primer. Chapter in Book: Trustworthy AI in Medical imaging, Elservier.

Cite Arxiv

(2023). An Intentional Forgetting-Driven Self-Healing Method For Deep Reinforcement Learning Systems. In ASE'23.

Cite Arxiv

(2023). Robustness assessment of hyperspectral image CNNs using metamorphic testing. In IST.

Cite Arxiv

(2022). SmOOD: Smoothness-based Out-of-Distribution Detection Approach for Surrogate Neural Networks in Aircraft Design. In ASE'22.

Cite Arxiv

Teaching

 
 
 
 
 
Content Creator & Instructor - MLOps Upskilling Program
Institut de valorisation des données (IVADO)
Oct 2023 – Feb 2024 Montreal
 
 
 
 
 
Teaching Assistant - System Design (INF5060)
Institute of Sciences, Technologies and Advanced Studies of Haiti
Jan 2022 – May 2022 Haiti
 
 
 
 
 
Teaching Assistant - Software Testing and Validation (LOG3430)
Polytechnic Montreal
Sep 2019 – Dec 2019 Montreal
 
 
 
 
 
Teaching Assistant - Advanced Software Testing (LOG6305)
Polytechnic Montreal
Jan 2019 – May 2019 Montreal

Training

Tensorflow Developer Certificate
See certificate
Building Conversational AI Applications
See certificate
Computer Vision for Industrial Inspection
See certificate
Building Transformer-Based Natural Language Processing Applications
See certificate
Reinforcement Learning Specialization
See certificate
Machine Learning Engineering for Production (MLOps)
See certificate
Designing and Implementing Data Science Solutions on Azure
See certificate
Developing SQL Databases
See certificate
Power Platform Developer
See certificate
Analyzing Data with Power BI
See certificate