Charles is currently unavailable
Charles Demontigny
Verified Expert in Engineering
Data Science Developer
Montreal, QC, Canada
Toptal member since February 11, 2022
Bio
Charles is a senior data scientist with 6+ years of experience in Python programming, SQL, predictive analytics, and data-driven marketing on the Google Cloud Platform. He's been working with a wide variety of clients, from startups to Fortune 500 companies. Charles has a good understanding of the business aspects behind the technical work, and he can deliver through the entire data pipeline process while analyzing large datasets using data science techniques and dashboards.
Portfolio
Large Consulting Firm
Python, Scikit-learn, Spark, Data Visualization, Snowflake, Amazon S3 (AWS S3)...
Campus Coach
Python 3, FastAPI, MongoDB, Docker
GoCoupons
Python 3, Optical Character Recognition (OCR), Entity Extraction...
Experience
- Pandas - 7 years
- Python - 7 years
- SQL - 5 years
- Scikit-learn - 5 years
- Google Cloud Platform (GCP) - 3 years
- Google BigQuery - 3 years
- Plotly - 2 years
- Python API - 1 year
Preferred Environment
Python, SQL, Pandas, Plotly, Dash, Google BigQuery, Scikit-learn, Google Cloud, Python API, Google Cloud Platform (GCP)
The most amazing...
...project I've worked on included helping a startup grow its user base with ML, predicting which users are most likely to convert to premium, and targeting them.
Work Experience
Data Science Engineer
2022 - 2023
Large Consulting Firm
- Integrated and standardized multiple data sets from various vendors into the Snowflake data warehouse, enhancing the data quality and accessibility for machine learning models used by the firm's case teams.
- Developed and maintained robust Airflow DAGs, ensuring daily data loads were efficient and accurate, resulting in improved database performance and reliability.
- Utilized a diverse tech stack including Python, Airflow, AWS S3, and Snowflake, effectively streamlining data engineering processes and delivering reliable, scalable solutions to support the firm's machine learning initiatives.
- Collaborated with cross-functional teams to identify data requirements and implement data engineering best practices, contributing to the successful completion of various projects and driving measurable value for the firm's case teams.
Technologies: Python, Scikit-learn, Spark, Data Visualization, Snowflake, Amazon S3 (AWS S3), Apache Airflow
Data Scientist
2021 - 2022
Campus Coach
- Collaborated with Campus Coach, a training app for runners, to develop targeted marketing campaigns by identifying and segmenting their free user base.
- Employed scikit-learn and Python to create a propensity score model, which helped predict the likelihood of users converting to paid subscriptions.
- Utilized the model's insights to assist Campus Coach in tailoring marketing efforts, resulting in more effective campaigns and increased user conversion rates.
Technologies: Python 3, FastAPI, MongoDB, Docker
Machine Learning Engineer
2021 - 2022
GoCoupons
- Developed a system to read and process grocery invoices for GoCoupons.ca, utilizing Google Cloud AI's Vision and Natural Language APIs with the Python SDK for product and banner detection.
- Integrated GPT-3.5 turbo via the OpenAI API to accurately extract every product from the invoice images, enhancing the overall data extraction process.
- Collaborated with the couponing company to implement this solution, resulting in a more efficient and automated product recognition system.
Technologies: Python 3, Optical Character Recognition (OCR), Entity Extraction, Data Extraction, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Google AI Platform, Google Cloud, OpenAI GPT-3 API, Artificial Intelligence (AI), Computer Vision, Text Recognition
Machine Learning Engineer
2019 - 2020
Equifax
- Designed and implemented a machine learning-based system to predict the real estate value of over 2 million properties across Canada.
- Utilized the XGBoost algorithm to build an accurate and efficient prediction model, significantly enhancing the property valuation process.
- Enabled data-driven decision-making for investors, property owners, and real estate professionals by providing reliable property value estimates.
Technologies: Python 3, XGBoost, Azure, Scikit-learn
Data Scientist
2019 - 2020
King & Partners
- Collaborated with a US-based hotel chain to identify and target high-value customers through the analysis of booking data and CRM records.
- Implemented CLV (Customer Lifetime Value) predictions and segmentation using Python and scikit-learn, enabling the hotel chain to focus on retaining their most valuable guests.
- Leveraged the insights gained from the analysis to inform marketing and customer service strategies, ultimately enhancing guest satisfaction and loyalty.
Technologies: Python, SQL, Machine Learning, Data Analytics, Data Science, Google Cloud Platform (GCP)
Data Scientist
2017 - 2019
JLR Solutions Foncières
- Developed, integrated, improved, and maintained a machine learning model able to estimate the market value of the houses in Canada with LightGBM in Python and SQL for the ETL process.
- Built a housing price index based on the three-stage least-square regression methodology by Case and Shiller (1987) using Python and SQL Oracle.
- Wrote reports on the state of the real estate market. Produced econometric analyzes based on real estate microdata. Communicated the analysis produced from the data and was interviewed on radio and newspapers about these studies.
Technologies: Python, Machine Learning, Data Analytics, Data Science, Google Cloud Platform (GCP)
Experience
Predicting Customer Lifetime Value
https://github.com/Charles-de-Montigny/predict_customer_lifetime_valueIn this project, transactional data is transformed into a single view per customer. Average order value, purchase frequency, and recency are then used to feed a probabilistic model to produce a prediction of future customer lifetime value.
The marketing department uses this to produce targeted campaigns on high-value customers, those at risk of churn, or high potential. Finally, it is possible to push lists of our best customers to Google Ads and Facebook Ads to acquire customers that look like our best.
In this particular case, open data is used. However, I developed this type of project with several clients, but I am not allowed to share it for obvious data privacy reasons.
The marketing department uses this to produce targeted campaigns on high-value customers, those at risk of churn, or high potential. Finally, it is possible to push lists of our best customers to Google Ads and Facebook Ads to acquire customers that look like our best.
In this particular case, open data is used. However, I developed this type of project with several clients, but I am not allowed to share it for obvious data privacy reasons.
SEO Espion
Co-founded a startup and built an MVP designed to assist SEO practitioners in working more efficiently and accurately through a data-driven approach.
Developed a system that scrapes Google search result pages for a given query and website URL, then extracts and analyzes the first 100 websites, creating ranking factors (features) from each HTML page.
Leveraged machine learning models using Python, FastAPI, Google Cloud Platform, Cloud Run, Firebase, BigQuery, and scikit-learn to understand feature importances and generate actionable roadmaps for optimizing target web pages, streamlining the SEO process for freelancers.
Developed a system that scrapes Google search result pages for a given query and website URL, then extracts and analyzes the first 100 websites, creating ranking factors (features) from each HTML page.
Leveraged machine learning models using Python, FastAPI, Google Cloud Platform, Cloud Run, Firebase, BigQuery, and scikit-learn to understand feature importances and generate actionable roadmaps for optimizing target web pages, streamlining the SEO process for freelancers.
Montreal Canadiens Dash App
Developed and deployed a Plotly-Dash app to analyze and compare past and current players of the Montreal Canadiens Hockey team, providing valuable insights for team performance evaluation.
Utilized Python, Plotly-Dash, Docker, and Heroku for app development and deployment, ensuring a smooth user experience and accessible interface.
Utilized Python, Plotly-Dash, Docker, and Heroku for app development and deployment, ensuring a smooth user experience and accessible interface.
Education
2015 - 2017
Master's Degree in Econometrics
ESG-UQAM - Montreal
2012 - 2015
Bachelor's Degree in Economics
ESG-UQAM - Montreal, Canada
Certifications
OCTOBER 2019 - PRESENT
DeepLearning.AI – Deep Learning Specialization
Coursera
Skills
Libraries/APIs
Pandas, Scikit-learn, Python API, XGBoost
Tools
Plotly, Apache Airflow, Google AI Platform, BigQuery
Languages
Python, Snowflake, SQL, Python 3
Storage
Amazon S3 (AWS S3), Google Cloud, MongoDB
Platforms
Google Cloud Platform (GCP), Docker, Azure
Frameworks
Spark
Other
Econometrics, Machine Learning, Data Analytics, Data Science, Dash, Google BigQuery, Customer Data, User Analysis, Deep Learning, Data Visualization, Optical Character Recognition (OCR), Entity Extraction, Data Extraction, Natural Language Processing (NLP), OpenAI GPT-3 API, FastAPI, Web Development, Web Dashboards, Artificial Intelligence (AI), Computer Vision, Text Recognition, Generative Pre-trained Transformers (GPT)
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