Scientist in Action

The IVADO “Scientist in Action” program aims at startups interested in hosting a master’s-level intern to advance an innovation project in artificial intelligence.

The master’s student participating in the program will complete a credited internship in addition to gaining entrepreneurial experience and exploring new career opportunities.

The program objective is to match talents to the entrepreneurial ecosystem. This is achieved by means of a credited applied research internship at a startup company.

The candidate acquires experience in the entrepreneurial environment and is in a better position to make career choices.

For the startup, the program offers a unique opportunity to access sought-after expertise at an affordable cost, while promoting its organizational culture and potentially recruiting a new team member.

Any project involving artificial intelligence and its applications, for example:

• Business intelligence
• Operations research
• Data science
• Data visualization
• NLP
• Machine learning
• Etc.

For examples of mandates, here is a summary of some internship offers from previous editions.

2024

Lisplogics – Real-Time Battery Switching and Broken Bike Pickup Algorithm for Bike-Sharing

Lisplogics is an operations research company focusing on the optimization and automation of real-time operations in the shared mobility and transportation sectors, with customers including Bixi and Communauto.

The project consists in developing a dynamic algorithm to optimize battery swapping and broken bike recovery operations in a bike-sharing system (1 algorithm for both). The aim is to improve the efficiency and responsiveness of the current system by integrating a predictive model and a dynamic routing algorithm.

Neuro Solutions Group – Optimizing an avatar parent coach for children with neurodevelopmental disorders.

Dive into the future of digital health with an internship at Neuro Solutions Group! We’re looking for an intern with a passion for AI to revolutionize the way parents of children with neurodevelopmental disorders interact with our avatars. You’ll develop an advanced chatbot for smooth communication, implement machine learning for nuanced responses, and create personalized, data-driven notifications for parental insights. Collaborate with experts to integrate a comprehensive parenting guide and improve tracking of child behavior and medication, to better manage treatment plans. Join us in shaping the future of child development and parenting support with cutting-edge technology!

Our flagship product, Kairos, aims to enrich the lives of children with neurodevelopmental disorders such as ADHD and ASD, and optimize their care by healthcare professionals. Integrating cutting-edge techniques such as gamification and artificial intelligence, Kairos offers immersive and effective therapy, enriched by key data to personalize treatments. Launched in Canada in 2022, Kairos quickly gained recognition, winning a Pediatric Device Innovation Award in Texas. Pilots are currently underway with CHUS Sherbrooke and CHU Ste-Justine. Our sights are now set on Europe, where Kairos is poised to become Canada’s first DiGA application, prescribed for children aged 6 to 12. This is an excellent opportunity to collaborate with experts in child development, parenting psychology and AI technology.

2023

Sport AI – Computer vision and natural language applied to soccer performance analysis.

At Sport AI, we use artificial intelligence to analyze the performance of soccer teams in real time to help the coaching team in their decision-making in addition to helping players improve.

The intern will have the opportunity to work on one of the following two requirements:

• Develop a computer vision model (supervised learning, OpenCV or TensorFlow framework) capable of detecting players and the ball in soccer videos in real time. In addition, a natural language recognition (NLP, PyTorch) model will be developed to label match events using voice commands.
• The detection model should be able to recognize people on the pitch and follow them throughout the video. The NLP model will have to result in a model that can be integrated with our performance analysis platform (operator) to communicate with our database in real time.

2022

Mely.AI – Development of an artificial intelligence model to facilitate data extraction from forms

At Mely.ai, we’re changing the supply chain and logistics landscape by providing the industry with AI-assisted tools. We want to offer an unprecedented experience to our users with the latest technologies to better serve their customers.

Mely.ai’s current added value is the ability to extract data from forms using specific templates. We also have a method for extracting data without a template, but it lacks precision.

The aim of the project is to create a generic model that will enable us to better extract data from a form into structured data. Using computer vision techniques, we will be able to identify tabular data as well as available key values.

As part of this internship, your responsibilities will include:

• Analyze the company’s current methods in order to make an assessment;
• Propose an alternative solution to the current information extraction method;
• Implement the solution in proof-of-concept mode;
• Deploy the solution on the Mely.ai SaaS platform.

Although the program is open to all students enrolled in Campus Montréal institutions (Université de Montréal, Polytechnique Montréal, HEC Montréal), McGill University and Laval University who meet the eligibility criteria and demonstrate the abilities required for the project, the main academic programs from which the students will be drawn are as follows:

HEC Montréal

• M.Sc. in Business Intelligence
• M.Sc. in Data Science and Business Analytics

Polytechnique Montréal

• Modular Master’s (or DESS) in Computer Engineering
• Engineering and Data Analytics option
• Artificial Intelligence in Interactive Digital Entertainment option

 Université de Montréal

• Master’s in Computer Science – Supervised Internship option
• DESS in Machine Learning
• Professional master’s degree in machine learning

Université McGill

• M.Sc. (Non-Thesis) Computer Science
• Master in Engineering and Computer Science

Université Laval

• Master in Computer Science – Artificial Intelligence
• M.Sc. Business Analytics

• November 28, 2024 to midnight January 6, 2025: Call for internship projects from startups
• January 7 to January 31, 2025: Projects selected and approved by partner universities
• February 3 to 23, 2025: Publication of internship offers and interviews
• February 3, 2025, 12pm to 1pm: Student event: “Find your summer internship in AI”. Selected startups present their internship projects
• February 24 to March 21, 2025: Finalization of internship agreements and submission of Mitacs applications
• From May 2025: Start of internships.

Scientist in Action is a program of credited summer internships in AI carried out by master’s students.

Ten startups will be selected and receive IVADO co-funding for one Mitacs internship unit. The official start date is subject to Mitacs approval. We are not responsible for these delays.

Option 1: Polytechnique, HEC, McGill and DESS Mila

Duration: 4 months full-time or 6 months part-time

Value: $15,000

Cost:

Startup: $3750 + tx
IVADO: $3750
Mitacs: $7500

Option 2: Université Laval and Université de Montréal (DIRO and Mila)

Duration: 6 months full-time or 8 months part-time

Value: $30,000

Cost:

Startup: $11250 + tx
IVADO: $3750
Mitacs: $15,000

The startup’s contribution will be invoiced by Mitacs and must be paid on receipt, prior to the start of the internship. Any delay in payment of the contribution will result in a postponement of the start of the internship.

For more information on the Mitacs Business Strategy Internship program.

For the student

• Be enrolled in an Master program at one of our partner universities: Université de Montréal, Polytechnique Montréal, HEC Montréal, McGill University, Université Laval;
• Register the internship as part of your course load (supervised project or credited internship);
  NB. It is the student’s responsibility to take the administrative steps to register the internship in his or her academic program.

• Submit an application following the home university’s standard process.

 For the startup

• Have less than 20 employees;
• Have a competent person to host the intern (the project does not rest on the shoulders of the intern);
• Be part of or have completed a program with a recognized entrepreneurial support organization (accelerator or incubator) within the last three years;
• Pay the invoice of $3750 + tx upon receipt to Mitacs before the start of the course.

Reminder: The internship will not be approved and cannot begin until the invoice has been paid. This is a determining factor in the realization of the internship and the start date. No exceptions are permitted.

For the student

The list of available internships and information about each application process will be available in the “List of available internships” section below.

For the startup

Complete and submit this form before midnight January 6, 2025. Any submission that is incomplete or sent by any means other than the form will be automatically rejected.

We suggest that you copy and paste your text once written.

If you have any questions, please contact: entrepreneuriat@ivado.ca

1. Bia Éducation: Optimizing scientific recommendation processes via the analysis of anonymized clinical data.

Bia Éducation is an innovative Quebec company specializing in healthcare e-learning and scientific knowledge transfer. Our mission is to provide healthcare professionals with fast, effective access to scientific research, enabling them to improve the quality of care they provide. We are currently developing an intelligent assistant based on artificial intelligence, designed to integrate scientific recommendations and alerts directly into Electronic Medical Records (EMRs).

The aim of the project is to develop a high-performance scientific recommendation system using advanced data science techniques. This system will analyze anonymized clinical data to: Determine the priority needs of healthcare professionals based on patient profiles. Optimize scientific alerts sent to clinicians, based on predictive analyses of recurrent clinical cases. Personalize recommendations according to clinical contexts and medical specialties.

For more details and to apply, see the full offer:

• HEC – Portail Ma Carrière, offer # 70091
• ULaval – Internships and jobs portal
• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer # 93856
• Université de Montréal / Mila – you have received an email to this effect

2. Cloud Solution Provider: AI Model Development for Cloud Cost and Carbon Footprint Management

Cloud Solution Provider is a cloud assistant that simplifies managing cloud costs and carbon footprint with intelligent insights centralized in one platform.

Join us to develop cutting-edge AI models in the Cloud FinOps domain.
During this internship, the objectives will be:

1. Developing an anomaly detection model for cloud cost data.
2. Designing a contextual forecasting model to predict future cloud costs.
3. Creating a cost simulation model for “what-if” scenarios to optimize resource allocation strategies.

Main tasks and responsibilities:

• Develop machine learning models using frameworks like PyTorch, TensorFlow, or Scikit-learn to detect anomalies in cloud cost data.
• Build a forecasting model to predict future cloud costs based on historical usage and contextual factors (e.g., scaling policies, services).
• Design and implement a simulation model to assess the impact of different pricing scenarios and strategies.

For more details and to apply, see the full offer:

• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer # 93838

3. Scopia Tech: Computer vision algorithm for laparoscopic surgical navigation

Scopia helps surgeons perform minimally invasive procedures under camera guidance (endoscope, laparoscope, etc.).  Our solution adds layers of intelligence onto camera images to improve visualization, navigation, diagnostic and intervention.

The internship is part of a project to develop laparoscopic navigation.  The goal of the project is to superimpose, in real time during surgery, hidden anatomy onto laparoscopic images of the abdomen.  This helps reduce the risks of surgical complications by avoiding unwanted injury to critical anatomy.  The patient’s 3D anatomy can be obtained from pre-operative scans and must be registered onto the 2D laparoscopic images in real time.  This involves creating algorithms to automatically solve several sub-problems using computer vision methods, such as 2D-3D registration of a deformable object, calculation of 3D surface from 2D images, computation of the camera position from 2D images of the environment, automatic recognition of anatomical structures on 3D images, semantic segmentation of 3D scans, etc.  The internship will focus on automatic recognition of 2D images.

The intern will suggest, implement and test artificial intelligence algorithms to solve the problem, including preparing training data, training the algorithm, and testing the inference.

For more details and to apply, see the full offer:

• ULaval – Internships and jobs portal
• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93835
• Université de Montréal / Mila- you have received an email to this effect

4. Uwear: Train a DiT (diffusion models with transformers) adapter for fashion visuals

Uwear.ai is a Montreal-based startup founded in March 2023 with the mission of building the best virtual fashion photographer. We have trained a model capable of generating fashion photos of products worn by virtual models, starting from a single photo of the unworn garment. Our tool allows small and medium businesses with limited budgets to access high-quality photoshoots and enables larger companies to save significant time and resources.

Several research teams worldwide are continually tackling this use case. However, none of the open-source models we tested were satisfactory for our users, which compelled us to develop our own solution with our proprietary dataset.

The goal of this project is to train the second version of our adapter on the new generation of image models, DiT (FLUX and SD3.5). This new generation is expected to bring greater realism and more faithfully reproduce garments with fine details, an area where we currently face challenges.

• Data Adjustment: We have a dataset of over 800,000 labeled and cleaned photos, but some architectures to be tested require additional transformations (e.g., extracting a person’s pose from an image).
• Technological Monitoring: Reading the latest research papers on the topic and extracting techniques and approaches that could be applied to our use case.
• Implementation of Selected Approaches: Validating their reliability by modifying training scripts or AI models.
• Model Training: Using PyTorch and CUDA for training models.

For more details and to apply, see the full offer:

• HEC – Portail Ma Carrière, offer #70094
• ULaval – Internships and jobs portal
• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93837
• Université de Montréal / Mila you have received an email to this effect

5. Trebble: Multimodal Disfluency Detection in Multi-turn Conversations

Trebble is building the next generation of AI-powered audio and video editing software for everyday creators. Our platform helps entrepreneurs, educators, and small business owners save time by automating tedious editing tasks while delivering professional-quality results. Designed for non-professionals, Trebble leverages advanced AI to simplify the editing process, making it accessible and efficient for users who lack the resources to hire professional editors. Our mission is to empower creators to focus on their message and audience while we handle the heavy lifting of editing.

One of the most common edits that almost everyone makes when preparing content for public sharing is removing disfluencies. Disfluencies include filler pauses such as “uhm” and “ahs,” repetitions, false starts, partial words, stutters, and more. Removing them manually using traditional editing tools is time-consuming. Trebble currently provides a partial solution to help with disfluency removal.

The objective of this project is to develop a multimodal disfluency detection mode.The selected student will build a system that uses both transcription and acoustic signals to detect disfluencies. Audio input will be processed through a fine-tuned acoustic model (such as WavLM) to generate acoustic representations of disfluencies. Transcriptions, generated by our commercial ASR system with word-level timestamps, will be input to a fine-tuned BERT model to create language representations of disfluencies. These representations will be upsampled to frame level and combined into a BLSTM fusion network for final detection. If time permits, the student will also explore a promising multimodal LLM-based approach for disfluency detection.

For more details and to apply, see the full offer:

• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93836

6. Paperminds Technology: Bilingual Voice Sentiment Analysis: Enhancing AI’s Emotional Intelligence

Paperminds is a conversational intelligence company that rethinks the way brands connect with their stakeholders through the art of conversation and the science of understanding. We combine behavioral science with advanced deep learning technologies to create hyper-personalized conversations at scale, transforming traditional transactions into authentic, lasting relationships.

Our solutions, including docking station and conversational research platforms, are already deployed with established brands. Our Privacy Shield Framework (PFS) ensures seamless collection and analysis of conversations to uncover unique insights into consumer emotions and behaviors. These insights help brands grow by placing the human experience at the heart of their operations.We train Big Language Models and text-to-speech technologies to create digital personalities capable of truly engaging, empathizing and connecting with users.

Project objectives and deliverables:

Technical Implementation:

• Functional prototype of voice sentiment analysis model supporting both French and English.
• Documentation of model architecture and implementation methodology.
• Performance metrics and evaluation reports for both languages.
• Integration code for existing platform compatibility.

Research & Analysis:

• Comparative analysis of different approaches to voice sentiment detection.
• Benchmark results and optimization recommendations.
• Technical documentation of chosen solutions and rationale.

Progress Tracking:

• Regular progress reports and presentations.
• Code repository with clear documentation.
• Final presentation of implemented solution and results.

For more details and to apply, see the full offer:

• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93834

7. Obius: Development of purchasing and cash flow planning algorithms.

Obius develops and markets AI-powered technology solutions to automate supply chain management processes.

With our software, Obius helps companies optimize their purchasing operations and avoid errors. After developing supply chain optimization software for e-commerce companies, Obius is extending its software offering to manufacturing companies.

The project involves digitizing and automating certain processes:

• Dynamic updating of configuration parameters
• Optimization of quality verification and data conformity processes
• Anticipation and planning of orders, automation of production workflows and product modification
• Production and performance monitoring, real-time operations management, work order and change request tracking
• Automated processing of drawings, BOM and parts list management

Main tasks of the trainee:

• Needs analysis and specifications
• AI model development and training
• Data processing and management
• Deployment and integration
• Collaboration and documentation
• Optimization and maintenance

For more details and to apply, see the full offer:

• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93832

8. CyberPNM: Integrating AI into a cybersecurity solution

CyberPNM offers a more human approach to tackling cybersecurity and organizational resilience through its suite of products and services: KENEP. This software as a service (SaaS), tailored to global markets, aims to strengthen the security of organizations and their employees by reducing the amount of personal information circulating about them on the internet. By monitoring digital footprints via OSINT tools, Kenep reduces the risks of exploitation while respecting users’ privacy. With the Vesta platform, employees receive personalized recommendations based on self-motivation principles, helping them to adopt better cybersecurity practices. The CYTO report supports users in the event of incidents, offering detailed investigation reports and support in the event of identity theft. Kenep also improves cyber governance and legal compliance, while fostering an accessible and proactive cybersecurity culture.

The aim of this internship is to pursue and develop the integration of an artificial intelligence solution within a cybersecurity platform. The intern will be responsible for implementing the various AI components, and improving detection, analysis and protection capabilities against cyber threats.

The intern’s main tasks and responsibilities:

• Develop AI modules for knowledge management, predictive analysis and safety rule validation;
• Implement data collection, processing and transmission mechanisms between the various AI modules;
• Design AI functionalities for local processing, management of sensitive data and generation of personalized alerts;
• Integrate a conversational interface to enable users to interact in real time in a mobile platform;
• Manage bias in AI algorithms.

For more details and to apply, see the full offer:

• HEC – Portail Ma Carrière, offre #70090
• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer  #93839
• Université de Montréal / Mila – you have received an email to this effect

9. Hivelighter: Automated RFP Research

Hivelighter is a SaaS provider specializing in personalized and perspective-based knowledge discovery and knowledge management. Hivelighter helps businesses in knowledge intensive industries (marketing, consultancies, law firms, educational organizations, etc) preserve human intelligence at the core of their competitive advantage

This project builds on existing features of Hivelighter’s platform. Specifically the robo-researching « Smart Collection » functionality, and our Read For Me™ automated perspective-based highlighting engine. The objective of this project is to automate the first, critical stage of RFP response: research.

In collaboration with the Hivelighter team, the intern will be required to:

1. Analyze and preprocess three years of comprehensive RFP-related data (from 100’s of RFPs — data includes the initial RFP, the « get smart » research decks generated, the relevant case studies used, the completed RFP responses, and outcomes).
2. Develop and fine-tune a custom LLM to extract key insights and queries from RFPs using zero-shot or few-shot approaches.
3. Implement algorithms to generate categorized research buckets based on new RFP input, user preferences and historical RFP performance.
4. Design and develop APIs for semantic AI search and integration with Hivelighter’s platform.
5. Collaborate on the development of the Read For Me™ profiles/perspectives (derived from RFP types, and key experts) required to deliver highly curated highlights and insights from internal and third-party sources.
6. Document methodologies, findings, and recommendations for deployment and scaling.

For more details and to apply, see the full offer:

• ULaval – Internships and jobs portal
• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93831
• Université de Montréal / Mila – you have received an email to this effect

10. Hint: Capitalize on the information gathered for the production of a prime-time media program

HINT specializes in the creation of relevant content. We support organizations by enabling them to harness the full potential of their data and expertise to create personalized, relevant and in-depth content.

As an intern, you’ll be at the heart of our mission, working on analyzing the needs of a major strategic and media client by contributing to the development of our intelligent search engine.

Project objective: To analyze how the knowledge gathered by researchers and journalists can be disseminated and used by the entire team/organization to create relevant content.

Main tasks, in collaboration with the HINT team:

• Collaborate with the HINT team and the customer to analyze existing knowledge flows within the customer’s organization.
• Contribute to the mapping of available data
• Participate in structuring knowledge flows to facilitate access and relevance of shared knowledge.
• Participate in the development of the startup’s search engine functionalities.
• Technology watch and research

For more details and to apply, see the full offer:

• McGill – you have received an email to this effect
• Polytechnique – La Ruche, offer #93840
• Université de Montréal / Mila – you have received an email to this effect

A committee composed of:

• Two to three IVADO technology transfer advisors, including the entrepreneurship advisor,
• The Equity, Diversity and Inclusion (EDI) Advisor

The committee will assess the applications and make a selection based primarily on compliance with eligibility criteria, the quality of the proposed internship project and the project’s impact on the company. The evaluation committee will be multidisciplinary and in line with EDI principles.

The IVADO entrepreneurship advisor will present the selected internship projects to the partner universities, who will judge which proposals meet the requirements of their programs and can be presented to their students.

As for the choice of interns, it’s the startup’s responsibility to interview and confirm the successful candidate to the entrepreneurship advisor before the date they’ll be given.

As this is not a research internship, all intellectual property belongs and remains with the startup. A confidentiality agreement (NDA) may be signed between the intern and the company. As this is not a research internship, all intellectual property belongs and remains with the startup. A confidentiality agreement (NDA) may be signed between the intern and the company.

For interns

Demonstrate integrity and respect in all aspects of your collaboration with the startup;

Respect the terms of the internship and ensure that objectives and deliverables are met to the best of your ability;

Mention IVADO in public communications about the internship and participate, whenever possible, in IVADO student activities.

For the startup company

Respect EDI principles when selecting candidates, in accordance with the IVADO EDI reference framework;

Consult the IVADO brochure on unconscious bias in recruitment;

Provide a working environment and infrastructure that enables the smooth running and completion of the project;

Demonstrate integrity and respect in all aspects of your collaboration with the intern;

Be available on February 9, from 10 a.m. to 12 p.m., for the presentation of internship projects to the student community;

Be available during the process leading up to the submission of the Mitacs application (a few meetings and e-mail exchanges are to be expected);

Complete the short feedback survey sent by IVADO at the end of the course;

Mention IVADO in public communications related to the intern’s work.

IVADO’s commitment to equity, diversity and inclusion

To ensure that all members of society benefit equally from the advancement of knowledge and opportunities, IVADO promotes the principles of equity, diversity and inclusion in all its programs. IVADO is committed to providing a recruitment process and research framework that are inclusive, non-discriminatory, open and transparent.