More than 90 projects completed since 2019

The globally-operating medical company has solved the challenge of monitoring patient movements on its high-performance beds with a solution based on artificial intelligence (AI) and technology.

By developing a motion detection tool using load cell and lidar measurements, as well as a real-time predictive model, the company has succeeded in integrating advanced functionality into its software interface. Managers can now benefit from efficient monitoring of patient movements, with predictions accurate to 80% for different classes of movement.

The company, which specializes in the placement of healthcare-related temporary staff, overcame the challenge of processing more than 1,000 weekly e-mails from various healthcare institutions. Using advanced technologies, including models based on Natural Language Recognition (NLP) and the GPT3.5 Large Language Model (LLM), the company developed a mailbox monitoring tool capable of interpreting data in e-mails and attached documents.

This solution simplified and standardized the staffing assessment process, extracting essential data and presenting it consistently to managers. By using Azure OCR to interpret images and PDFs, the company achieved a significant increase in productivity by eliminating request assessment time, improved responsiveness to incomplete requests, and enhanced customer service through fast and efficient request processing.

The company managing a B2C transactional platform has risen to the challenge of intense competition by developing solutions based on artificial intelligence (AI) to optimize the customer experience and simplify the data entry process. Faced with the diversity of merchants' classification methods, AI tools were deployed to suggest product categories based on names and descriptions, while an image recognition tool enhanced category prediction, also ensuring the detection of unauthorized items.

The results include better and more consistent data, accurate category prediction in over 80% of cases, and a significant reduction in the time needed to integrate new merchants into the platform.

The company, which develops software solutions to optimize road construction activities in Quebec, has taken up the complex challenge of rapid decision-making on worksites. Faced with weather variations impacting paving revenues and the complex legal environment surrounding road construction, the company has integrated intelligent dashboards into its software solutions.

These dashboards consolidated data from a variety of sources, including GPS-equipped sensors on worksites. Results include improved information sharing between industry stakeholders, more relevant decision-support tools thanks to real-time data, and increased sales as a result of this new software functionality.

The company, operating a digital platform focused on information for growing companies, took on the challenge of providing its customers, mainly B2B investors, with up-to-date and comprehensive information. Prior to our intervention, manual updating of the database involved consulting various internal documents and external sources.

The solution developed analyzed these documents, collected any missing information, and automatically integrated it into the database. Using generative AI models, the results of customer queries are presented in consolidated natural language reports. The results include automatically updated, higher-quality data, enhancing the customer experience with accurate, easy-to-read reports.

The company specializing in the resale of used vehicles overcame the complex challenge of standardizing available vehicle information from multiple sources to facilitate decision-making. With between 3 and 4 million vehicles on the market each year, the variety of suppliers' classification systems made data inconsistent. In addition, customers had their own criteria and vocabulary for assessing vehicle potential.

The solution involved the use of Natural Language Recognition (NLP) and Large Language Model (LLM) based models such as GPT3.5 to collect and standardize vehicle information, associating it with the customer's internal criteria. The results include an enhanced customer experience thanks to a standardized presentation of data, enabling effective comparison of vehicles whatever their source, as well as more informed decision-making thanks to the presentation of data in accordance with the user's specific needs in terms of attributes, vocabulary and classification.

A company specializing in the mining sector took on the crucial challenge of detecting anomalies in the excavation and crushing process, particularly as regards the operation of the crusher, an essential element in the mining operations chain. Existing manual controls based on granulometry were insufficient, and repairing or replacing the crusher was costly.

The solution involved using computer vision technologies to develop an artificial intelligence (AI) model capable of analyzing the size of each rock as it left the crusher. In the event of non-conformity with expectations, the system automatically reported an anomaly to the people in charge. The results included a significant improvement in the ability to detect anomalies, offering a potential reduction in equipment maintenance and repair costs through automated problem identification.

The recycling specialist overcame the challenge of fluctuating prices on international markets by improving its prediction of supply and demand. As price instability has a significant impact on profitability, the company has integrated machine learning models into its existing technological infrastructure. These models, based on analysis of international material price trends, transactions reported in public registers and internal history, have enabled more accurate anticipation of supply (suppliers) and demand (customers) trends. In addition, dashboards have been developed to optimize the use of available data and inform management decisions.

Results include a two-month forecast horizon for global price fluctuations, significant gains in productivity and competitiveness, leading to an increase in company value and subsequent acquisition. The model is able to anticipate any drop in demand in excess of 30%, enabling the customer to more effectively direct the actions of his sales representatives and ensuring greater stability in the flow of traded materials.