Open to opportunities

Francois David

@francoisdavid

Message

I build production machine learning and LLM automation systems for underwriting, biometrics, and fraud prevention.

Canada

Message

What I'm looking for

I'm seeking challenging, impactful roles where I can deepen my machine learning expertise, solve complex problems with collaborative teams, and contribute to computer vision, LLM, deep learning, generative-model, finance, or data science applications.

I'm building ML and LLM-powered automation at nesto for mortgage underwriting, including RAG tools, OCR document intelligence, and auditable rule-based validation systems.

At Jumio, I built production biometric systems for facial recognition, age estimation, face detection, and fraud prevention. My facial matching work reduced vendor dependency by 80% and cut annual costs by more than $500,000, while face detection improvements reduced costs by more than $400,000 annually.

I've also developed anonymization pipelines using StyleGAN2, published age-estimation research accepted at IEEE FG 2025, and worked on forecasting, adversarial training, and mobile applications. I'm especially interested in computer vision, LLMs, deep learning, generative models, finance applications, and data science.

Experience

Work history, roles, and key accomplishments

Current

Senior Machine Learning Developer

Current

Nesto

Feb 2026 - Present (7 months)

Developing ML and LLM-powered automation systems for mortgage underwriting, including RAG-based tools and OCR pipelines for document validation. Implementing explainable validation graphs and rule-based decision systems to support auditable underwriting outcomes.

Machine Learning LLM RAG OCR Document Intelligence Python PyTorch TensorFlow Keras scikit learn Pandas NumPy SQL Spark

Machine Learning Engineer

Jumio Corporation

Dec 2021 - Feb 2026 (4 years 2 months)

Specialized in face recognition, face detection, age estimation, and homography detection. Designed and implemented production-ready modules, including a facial recognition module that reduced vendor dependency by 80% and cut costs by over $500,000 annually, and an age prediction module evaluated by NIST as top-performing.

Face Recognition Face Detection Age Estimation Homograph Detection Biometrics Machine Learning Deep Learning Python PyTorch TensorFlow Keras scikit learn Pandas NumPy SQL Spark

Machine Learning Engineer Intern

Jumio Corporation

May 2021 - Dec 2021 (7 months)

Designed and implemented an anonymization pipeline integrating face detection, keypoint estimation, and synthetic face generation using GANs (StyleGAN2) for secure data augmentation.

Anonymization Face Detection Keypoint Estimation Synthetic Face Generation Gan StyleGAN Python PyTorch TensorFlow Keras scikit learn Pandas NumPy

Education

Degrees, certifications, and relevant coursework

Montreal Institute of Learning Algorithms (MILA)

Masters, Deep Learning

2020 - 2021

Grade: 4.23

Masters in Deep Learning from the Montreal Institute of Learning Algorithms (MILA) in association with University of Montréal, with a GPA of 4.23.

Concordia University

Bachelor of Computer Science, Computer Science

2017 - 2019

Grade: 3.94

Activities and societies: Member of the Dean’s List on all possible occasions.

Bachelor of Computer Science with a Minor in Mathematics & Statistics from Concordia University, with a GPA of 3.94 and Dean's List recognition.

HEC Montreal

Bachelor of Administration, Administration (Economics Orientation)

2015 - 2017

Grade: 3.80

Activities and societies: International Exchange Semester at Reykjavik University in Iceland in 2017; 3 Honor Mention Semesters on 3 possibilities.

Completed 60/90 credits of a Bachelor of Administration (Economics Orientation) at HEC Montreal, with a GPA of 3.80 and an international exchange semester at Reykjavik University in Iceland.