ML engineer focused on bringing models from prototype to production

I build ML systems that run in the real world: inference pipelines at scale, edge AI for safety-critical environments, and products used daily by major fleets across North America. Currently at Isaac Instruments.

Video Analytics Pipeline at Scale
Isaac Instruments, 2024 - present
Machine Learning Engineer
Designed and shipped an end-to-end video analytics service from prototype to production, processing fleet dashcam footage through 20 ML models. Collaborated with DevOps on monitoring, scheduling, and automated recovery. The system runs continuously across major Canadian and US trucking fleets.
7K+
Videos / hour
20
ML models
50%
Latency reduced
150K+
Daily requests
PyTorchONNXFastAPIDockerGrafanaAzure
Voice Assistant for Truck Drivers
Isaac Instruments, 2024 - present
Machine Learning Engineer
Built a hands-free voice-command system for truck drivers on resource-constrained edge hardware. Used TTS, STT, and OpenWakeWord to enable safety-critical interactions while driving, without relying on cloud connectivity.
TTS/STTOpenWakeWordEdge InferencePython
Document Scanning & Extraction
Isaac Instruments, 2024 - present
Machine Learning Engineer
Identified a gap in the existing solution, partnered with the product team to build a POC, and drove it to production. The system enhances tablet-based document scanning and data extraction, shipped across the entire client base.
Computer VisionOCRFastAPINext.js
DocTrack.io
Co-Founder, 2025 - present
Co-Founder & ML Engineer
Building a B2B SaaS platform for Quebec property management companies. Automates document classification, renaming, and expiry tracking through LLM-based email ingestion and extraction pipelines.
LLMsCloudflare WorkersFastAPIPostgreSQLTypeScript
Veterinary AI Assistant
Queva Inc., 2022 - 2024
Co-Founder & Chief AI Officer
Architected one of the first veterinary-assistance conversational agents powered by RAG. Built the IoT backend connecting database, mobile app, and ML inference. Trained a predictive model on IMU data to classify dog activities at 97.5% accuracy. Led a team of 6 and secured $250K from VCs.
97.5%
Classification accuracy
$250K
Funding raised
6
Team members
RAGLangChainAzureMongoDBFastAPIIoT
Low-Resource Language Translation
Personal, 2022 - 2025
Researcher & Developer
Fine-tuned an LLM with LoRA to translate English into Zarma/Songhai, my native language unsupported by Google Translate. Built both a word-vector approach and a full LLM-based system. Published across three technical publications.
LoRAHugging FaceWord EmbeddingsLow-resource NLP
Deploying PyTorch with FastAPI to Handle 120K+ Requests a Day
Data Science Collective
We Switched from TensorFlow to PyTorch and This Happened
Level Up Coding
When Google Translate Doesn't Support Your Language, You Build Your Own
Data Science Collective
How I Used AI to Create a Word Translator for My Native Language
Towards AI
Smile Detection for Image Moderation
Decathlon Digital

I'm an ML engineer with a computer engineering and electronics background, based in Montreal. I hold a Master's in Computer Engineering with a specialization in AI from the Université de Sherbrooke, where I graduated with an academic excellence scholarship.

My work sits at the intersection of machine learning and production engineering. I care most about the gap between a model that works in a notebook and a system that runs reliably at scale. That's where I spend my time.

Before Isaac, I co-founded Queva where I built the AI stack from scratch, raised VC funding, and led a small team. I've also done research in autonomous vehicles at the Université de Sherbrooke and computer vision work at Decathlon and CRIM.

Outside of work, I'm building DocTrack.io, writing about ML engineering on Medium, and contributing to low-resource NLP for African languages.

Currently

  • ML Engineer at Isaac Instruments
  • Co-Founder at DocTrack.io

Education

  • M.Eng. Computer Engineering, AI
  • B.Eng. Computer Engineering, Co-op
  • Université de Sherbrooke

Core Stack

PyTorch · ONNX · FastAPI · LangChain · Docker · AWS · Azure · PostgreSQL · Redis · gRPC · Hugging Face · MLflow · Grafana · GitHub Actions

Languages

  • French (native)
  • English (fluent)

Let's talk

Passionate about learning and always curious about new challenges. I'm happy to collaborate, exchange ideas, or simply have a conversation about ML, engineering, or anything in between.