MScAC Data Scientist Interns

We are always looking for amazing talent who can contribute to our growth and deliver results Geotab is seeking 3 Data Scientist Interns from the MScAC program at UofT who can help us discover the value out of our rich and large datasets. Our ideal team member will have the mathematical and statistical expertise you'd expect, but a natural curiosity and creative mind that's not so easy to find. As you mine, interpret and clean our data, we will rely on you to ask questions, connect the dots, and uncover opportunities that lie hidden within—all with the ultimate goal of realizing the data's full potential. You will join a team of data specialists, but will "slice and dice" data using your own methods, creating new visions for the future. If you love technology and are keen to join an industry leader — we would love to hear from you

As a Data Scientist Intern, you will be a crucial part of producing innovative analytics from Geotab's big data environment. We believe that there is collective value in the massive amount of data arising from all of our sensors across the globe. And most importantly, we believe this data can be used to improve safety, infrastructure, and productivity for our customers. You will work with our team of data scientists and engineers who have curated our data repository and have created a series of near real-time and historical datasets. You will leverage data mining techniques, Machine Learning, Deep Learning, Neural Networks and other Artificial Intelligence and statistical analysis tools to build models and high-quality prediction systems to help our customers make informed and smarter decisions. You will work closely with Data Analysts, Data Engineers, Professional Services, Data Privacy & Governance, firmware as well as other internal departments such as Software Development and Solutions Engineering to achieve these objectives.

• Please note that this posting is for students from UofT's MScAC program only.
• 8 month work-term beginning May 2026.
• Full-time, paid internship: Monday - Friday, 37.5hrs/week.
• Your first week at Geotab begins with 'GEO Launch' - a one-week Employee Orientation. Click here to learn more
• Learn more about the Geotab Campus Program here.
• This posting is for an existing vacancy.

Geotab Ace, Advance GenAI for Fleet Intelligence

• The primary objective of this internship is to develop a "Reasoning & Automation" solution that evolves Geotab Ace into a fully integrated, context-aware assistant. Key goals include:
  • Multi-Modal Ingestion: Implement a multi-modal ingestion layer (leveraging technologies like Google's LiveAPI) capable of processing screen content, audio, and text simultaneously to establish real-time user context.
  • Contextual Enrichment: Develop mechanisms to automatically enrich user prompts with environmental state (e.g., current page URL, visible data charts), eliminating the need for users to manually explain their context.
  • Proactive Assistance & Navigation: Extend the reasoning capabilities of the agent to not only answer data queries but also guide users through complex workflows ("Where do I go to...") and offer proactive support.
  • Optimization of RAG & SQL Generation: Continue to refine the underlying RAG architecture and "text-to-SQL" generation for Google BigQuery to ensure the answers provided are accurate, safe, and efficient.

Vehicle Collision Severity Prediction

• Building on foundational work completed in 2025, the 2026 project will focus on model development, evaluation and possibly implementation. The primary objective is to develop a baseline collision severity prediction model that moves beyond our current major/minor classification based on g-force magnitude. Success means creating a model that predicts business-relevant outcomes such as damage severity categories, estimated dollar amounts for repairs, or liability classifications. The intern will leverage cleaned datasets from 2025 (including claims data, public crash data, and other available sources) to engineer features and build initial models. Key deliverables include:
  • A baseline severity prediction model with documented performance metrics
  • Comparison of different modeling approaches and feature sets
  • Analysis of which data sources and features are most predictive of useful severity metrics
  • Recommendations for model refinement and production deployment.

Automated, Scalable Load Weight Detection in Heavy Duty Vehicles

• The project requirements and objectives are such that the end result is usable for the future work described. We should be able to detect when a vehicle on supported platforms is unloaded with a high degree of certainty. Additional bucketing of loaded vehicles into quartiles of maximum weighted load is a secondary goal. Once determining a minimum amount of driving data required to make a determination, all trips meeting that criteria should be categorized automatically with the development of a daily pipeline.

• Interact with Geotab's Big Data infrastructure on Google BigQuery using Python and SQL.
• Process, cleanse, and verify the integrity of data used for analysis.
• Select features and build and optimize classifiers using machine learning techniques.
• Use machine learning packages like scikit-learn and TensorFlow to develop ML models, features for ML models, and visualize features.
• Interface with product managers, data engineers, and software developers to gather requirements.
• Own the design, development, and maintenance of ongoing metrics, reports, analysis, dashboards, etc., and deliver complete analytics/reporting solutions to drive key business decisions.
• Use data mining, model building, and other analytical techniques to develop and maintain customer segmentation and predictive models.
• Make recommendations for new metrics, techniques, and strategies to improve the Geotab product suite.
• Support a platform providing ad-hoc and automated access to large datasets.

• Graduate or Postgraduate Degree in Computer Science, Software/Electronics/Electrical Engineering, or any related field.
• We welcome applicants from a range of experience levels with an excellent understanding of Python and SQL.
• Experience with parallelization tools like Apache Beam and Spark, and scheduling tools like Airflow is nice to have.
• Advanced pattern recognition and predictive modeling or Graph Data Modeling experience is a bonus.
• Experience with developing machine learning models and in feature engineering is a plus.
• Experience working with Google Compute Engine and Google BigQuery is nice to have.