Hi, I'm @CrSamson,

Applied AI Engineer

Trained on my bio, work, and projects

Selected work

Projects across AI agents, ML, and specialized AI.

13M+calls monitored via LLM-as-a-Judge1M+data points processed by OCR/RPA bot$30M+in annual savings supported30%fewer production incidents after refactor

AI Agents

StaffBot

AI-powered candidate-to-role matching engine.

LLM-driven resume + job description parsing
Semantic similarity scoring with embeddings
Multi-criteria weighted ranking system

PythonLangChainOpenAIEmbeddings

Brevio AI

A daily AI news digest that runs itself.

Multi-source ingestion (RSS, arXiv, HF, YouTube)
GPT-4o summarization with per-kind prompts
GitHub Actions cron + Neon Postgres + SMTP

PythonGPT-4oNeon PostgresGitHub Actions

Specialized AI

PharmaOCR

OCR engine that reads and understands pharmaceutical documents.

Layout-aware text extraction
Domain-specific entity recognition
Variability and edge-case test suite

PythonOCRComputer Vision

Splendor.ai

Neural networks that learned to play Splendor.

ISMCTS expert data generation
Imitation-learning neural networks
100K+ state-action training pairs

PyTorchNumPyGame AI

Data Science & ML

FTR Prophet

Walk-forward ML for energy market trading.

Walk-forward validation with anti-leak guards
36-feature LightGBM ensemble
$5.3M simulated net profit

PythonLightGBMpandasscikit-learn

CardioPredict

Comparative ML benchmark for cardiac disease risk.

Six-model comparative benchmark
Cross-validated performance metrics
Clinical-feature interpretability analysis

Pythonscikit-learnXGBoostJupyter

MaintenanceIQ

Profit-aware predictive maintenance.

Random Forest classifier
Profit-optimized decision threshold
Real industrial dataset

RrandomForesttidyverse

Monte Carlo Simulation

Decision trees vs. random forests across 8 scenarios.

8 scenarios × N MC iterations
Accuracy + runtime per model per run
Statistical comparison across scenarios

RrpartrandomForest

About

Industrial engineering background. Applied AI on production systems.

Currently

Shipping RAG and LLM-evaluation systems in production
Finishing my M.Sc. in Data Science at HEC Montréal (2026)
Open to discussing applied AI roles, freelance work, and interesting collaborations

Applied AI Engineer based in Montréal. I build production-grade machine learning and generative AI systems in an enterprise context, focused on the work that has to run after the demo ends. My background sits at the intersection of industrial engineering and data science, picked up at Polytechnique Montréal and HEC Montréal.

What I do today

Currently working on RAG pipelines, automated LLM-as-a-Judge evaluation frameworks, and performance monitoring systems for multiclass ML models in production. Before this, two years of RPA automation work taught me how to design systems that run reliably without supervision.

Outside of work

I build AI side projects across a few categories: agentic systems (StaffBot, Brevio AI), trading models (an HEC Data Challenge entry that generated $5.3M in simulated net profit), and specialized tools (PharmaOCR, Splendor.ai).

I co-founded the Polytechnique Business Case Center during my undergrad. Turning ambiguous business problems into structured analyses is most of the actual job.

From IE to applied AI

Industrial engineering taught me to look at systems end to end: flows, bottlenecks, what creates value and what just creates motion. Applied AI is the same lens with sharper tools. When a model takes hours of grinding work off someone's plate, the win is human, not technical.

Stack

AI & GenAI

RAG · LangChain · LLMs · Prompt Engineering · OpenAI API · LLM-as-a-Judge

ML & Data

PyTorch · TensorFlow · scikit-learn · XGBoost · LightGBM · pandas · PySpark

Infrastructure

Azure ML · Docker · MLflow · CI/CD · Git · PostgreSQL

Languages

Python · SQL · R · Java · English · French

Education

M.Sc. in Data Science

HEC Montréal

Winner projects across multiple hackathons

B.Eng. in Industrial Engineering

Polytechnique Montréal

Co-founder, Polytechnique Business Case Center