Julien St-Pierre Fortin, Developer in Montreal, QC, Canada
Julien is available for hire
Hire Julien

Julien St-Pierre Fortin

Verified Expert  in Engineering

Machine Learning Developer

Location
Montreal, QC, Canada
Toptal Member Since
November 13, 2019

Julien's expertise is deploying software solutions in data and AI. He is passionate about helping clients create innovative products and gain valuable insights from data. He genuinely cares about his clients and goes above and beyond to ensure highly successful projects.

Portfolio

Succession
Next.js, React, Terraform, AWS CLI, Git, OpenAI GPT-4 API, OpenAI GPT-3 API...
Coveo
C#, MySQL, .NET, Object-oriented Programming (OOP), Snowflake...
Mino Games
SQL, Scikit-learn, Pandas, Python, Data Science, Git, NumPy, BigQuery

Experience

Availability

Part-time

Preferred Environment

Slack, GitHub, Google Docs, Miro, Figma

The most amazing...

...product I've built is Succession HQ, a SaaS app to optimize operations for manufacturing SMBs.

Work Experience

CTO

2023 - PRESENT
Succession
  • Built retrieval-augmented generation algorithms to power the product's business insights engine.
  • Architected and built the data infrastructure, as well as the app's back end and front end collaborating with the engineering team.
  • Worked with the CEO to define business requirements and processes to reach our goals.
  • Collaborated with the design lead to build our design system.
Technologies: Next.js, React, Terraform, AWS CLI, Git, OpenAI GPT-4 API, OpenAI GPT-3 API, TypeScript, SQL

Software Engineer

2019 - PRESENT
Coveo
  • Played a role in developing Coveo's latest generative question-answering capabilities.
  • Unified data modeling for the service and support line of business.
  • Improved data access security of search engine data.
Technologies: C#, MySQL, .NET, Object-oriented Programming (OOP), Snowflake, Data Build Tool (dbt), AWS CLI, Terraform, Java, Scala, PyTorch, OpenAI GPT-3 API, Git, OpenAI GPT-4 API, TypeScript, NumPy, TensorFlow, Test-driven Development (TDD), SQL

Data Scientist

2019 - 2019
Mino Games
  • Built machine learning models from cohort KPIs to get early feedback on marketing activity, such as the likelihood of acquired cohorts' profitability.
  • Developed revenue forecast models using curve-fitting methods.
  • Built dashboards to monitor user behavior with SQL, BigQuery, and Mode Analytics.
  • Developed a Python library to streamline the machine-learning process.
Technologies: SQL, Scikit-learn, Pandas, Python, Data Science, Git, NumPy, BigQuery

Data Scientist

2016 - 2019
Gameloft
  • Constructed machine learning models using scikit-learn and TensorFlow based on user behavior to predict cohorts' lifetime value (LTV) and profitability.
  • Co-developed many data science tools, a model factory, and APIs with Pandas and Flask.
  • Built an automated task scheduler using Apache Airflow.
  • Researched user dynamics spanning multiple products with Markov chains and recurrent neural networks.
  • Co-developed a Vue-based user interface to visualize model predictions.
Technologies: Docker, Jupyter Notebook, TensorFlow, Matplotlib, Scikit-learn, Pandas, Python, Data Science, Git, PostgreSQL, NumPy, Apache Airflow

Research Assistant

2014 - 2015
INRS
  • Developed deep learning models with MATLAB to predict water temperature from meteorological data to support engineering at the Beauharnois hydropower plant in Quebec.
  • Conducted research on the development of a network for measuring and collecting water temperature data from covariates. We used mutual information and dimensionality reduction methods using MATLAB.
  • Conducted web scraping of the Statistics Canada website to obtain meteorological data using Python.
Technologies: Deep Learning, MATLAB, Python, Data Science

Research Intern

2013 - 2013
TRIUMF
  • Participated in a research internship at TRIUMF on the IRIS experiment with Dr. Rituparna Kanungo.
  • Improved detector calibration with elastic scattering simulations in C++. The goal of the experiment was to demonstrate the particular halo structure of the lithium 11 isotope.
  • Collaborated with an international team of physicists.
Technologies: C++, Data Science

Learning Generated Abstractions

https://julienstpierrefortin.com/posts/learning-abstractions/
When we program an AI to learn simple yet purely virtual representations, what happens in the middle of the process? I used Generative Adversarial Networks (GANs) to generate abstract images from samples obtained via simple procedures in Python.

Reliable Vision System for Wildlife

The goal of this project was to build an autonomous camera system that automatically detects and counts bird species based on input images. We used YOLOv3 and Inception ResNet v2 algorithms for bird detection and classification, respectively.

Installation with Ed Fornieles Studios

https://capesaro.visitmuve.it/en/mostre-en/archivio-mostre-en/breathless-london-art-now/2019/10/20942/breathless/
I built a Node.js back end to power a Unity art installation at Ca' Pesaro Venice art gallery during La Biennale di Venezia 2019. It was connected to real-time financial analytics, running in a Docker container on AWS infrastructure.

Languages

Python, C#, SQL, Snowflake, Java, TypeScript, C++, Scala

Libraries/APIs

Scikit-learn, NumPy, Pandas, TensorFlow, Matplotlib, PyTorch, React, Keras, Node.js

Tools

Git, AWS CLI, Terraform, MATLAB, BigQuery, Apache Airflow, Slack, GitHub, Google Docs, Figma

Paradigms

Data Science, Test-driven Development (TDD), Object-oriented Programming (OOP)

Platforms

Jupyter Notebook, Docker

Other

Deep Learning, Artificial Intelligence (AI), Machine Learning, Modeling, Physics, OpenAI GPT-3 API, Numerical Methods, Data Build Tool (dbt), OpenAI GPT-4 API, Generative Adversarial Networks (GANs), Miro

Frameworks

.NET, Next.js, Unity

Storage

MySQL, PostgreSQL

2015 - 2018

Master of Science Degree in Data Science and Operations Research

HEC Montréal - Montreal, Quebec, Canada

2011 - 2014

Bachelor of Science Degree in Theoretical Physics

Université Laval - Québec, Canada

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring