Senior Data Scientist | Scientifique de données senior

Jesta I.S.•Westmount, QC

1d•Hybrid

About The Position

We are seeking a Senior Data Scientist with deep expertise across the full spectrum of modern AI — from classical predictive modelling to large language models, agentic AI systems, and computer vision. This is a senior individual contributor role requiring strong autonomous judgment, technical leadership, and the ability to define and execute complex AI workstreams with minimal supervision. The successful candidate will take full ownership of our existing production AI stack — maintaining, improving, and scaling models and pipelines already in active client use — while contributing meaningfully to the development of new capabilities. You will work at the intersection of predictive modelling, generative AI, and enterprise software, with your work deployed against real client data at scale.

Requirements

PhD in Artificial Intelligence, Data Science, Computer Science, or a related quantitative field strongly preferred.
Candidates with a Master’s degree and exceptional industry experience will also be considered.
Minimum 5 years of industry experience in applied data science (excluding academic research, internships, and research positions).
Proven experience deploying and maintaining production-scale ML systems.
Experience owning existing AI platforms as well as developing new capabilities.
Hands-on expertise with LLMs and Generative AI, including RAG, prompt engineering, structured outputs, and evaluation frameworks.
Experience building agentic AI systems with tool use, orchestration, and multi-step reasoning.
Working knowledge of MCP or comparable agent orchestration frameworks.
Experience developing and deploying computer vision models.
Demonstrated ability to independently lead complex technical initiatives.
Advanced Python, including PyTorch, scikit-learn, LightGBM, XGBoost, TensorFlow/Keras, pandas, polars, NumPy, and LLM frameworks (LangChain, LlamaIndex, Anthropic/OpenAI SDKs); R is an asset.
Forecasting libraries including Prophet, DeepAR, ARIMA/SARIMA, Croston/TSB, and ensemble methods.
Experience with Kedro or similar workflow orchestration frameworks and large-scale batch processing.
MLOps expertise including MLflow, Docker, CI/CD, Git, experiment tracking, model registries, drift detection, and retraining strategies.
Experience with AWS (SageMaker, Lambda, S3, IAM, Batch), Azure ML, Azure DevOps, and Snowflake/Snowpark.
REST API development and integration using frameworks such as FastAPI or Flask.
Strong statistical foundations including time series analysis, Bayesian methods, causal inference, experimental design, and hypothesis testing.
Advanced SQL for enterprise-scale analytics.
Highly autonomous with the ability to independently scope and deliver complex initiatives.
Excellent written and verbal communication skills with technical and executive audiences.
Strong analytical rigor and sound technical judgment.
Collaborative mindset with cross-functional engineering and MLOps teams.
Comfortable making decisions and driving outcomes in ambiguous environments.

Nice To Haves

Experience in commercial demand forecasting or time series modelling is a strong asset.
Experience with signal processing techniques is an asset.
Experience with ERP systems, retail data models, or supply chain data is an asset.

Responsibilities

Design, develop, and optimize machine learning models for demand forecasting, inventory optimization, and pricing analytics.
Perform exploratory data analysis (EDA), model diagnostics, and data quality assessments.
Engineer advanced features including promotions, seasonality, holidays, stockouts, lag variables, external signals, and signal processing techniques.
Design and execute model experiments, hypothesis testing, oracle testing, and statistical evaluations.
Evaluate and benchmark forecasting approaches including LightGBM, Random Forest, Gradient Boosting, Deep Learning (PyTorch), DeepAR, ARIMA/SARIMA, Prophet, ensemble methods, and Croston/TSB.
Own the full model lifecycle from development and backtesting to deployment, monitoring, drift detection, and retraining.
Design and implement LLM-powered applications using RAG, fine-tuning, prompt engineering, structured outputs, and vector databases.
Build AI systems that securely interact with enterprise data through governed APIs.
Evaluate and integrate commercial and open-source foundation models into production.
Develop explainability and transparency mechanisms for enterprise AI solutions.
Design and develop autonomous AI agents with multi-step reasoning and tool use.
Build integrations using Model Context Protocol (MCP) and tool-calling architectures for ERP data access.
Implement human-in-the-loop (HITL) workflows, role-based security, and approval mechanisms.
Establish standards for reliability, traceability, and auditability of agentic systems.
Design Semantic Read APIs connecting AI models to ERP data securely and reliably.
Build scalable batch inference and feature pipelines on AWS SageMaker and Azure ML.
Contribute to CI/CD automation, model validation, and deployment pipelines.
Collaborate with Data Engineering and MLOps teams on Dockerized deployments, APIs, monitoring, and scalable inference.
Define standards for feature stores, data pipelines, and model versioning.
Partner with IT and Security teams to ensure compliant AI deployments.
Develop and deploy computer vision solutions for product classification, visual merchandising, and image-based retail applications.
Integrate multimodal vision capabilities into forecasting and AI agent workflows.
Evaluate and apply modern computer vision architectures to production use cases.
Own and maintain production AI models and pipelines while driving continuous optimization.
Balance maintenance of existing solutions with development of new AI capabilities.
Identify and remediate technical debt, performance bottlenecks, and scalability issues.
Serve as the technical owner for AI model performance and production incident resolution.
Maintain comprehensive documentation for models, pipelines, and experiments.
Ensure reproducibility through experiment tracking, Git version control, and model lineage.
Share knowledge through code reviews, technical documentation, and internal presentations.
Contribute to R&D documentation and formal technical specifications.
Independently scope, design, and deliver AI solutions with minimal supervision.
Mentor team members through code reviews and architectural guidance.
Translate business problems into robust AI solutions and communicate results to technical and business stakeholders.
Apply responsible AI principles, including fairness, transparency, and model risk management.