Artificial intelligence

One of our clients, a leading pharmaceutical company, was interested in exploring the potential of external patient data in the United States and in Germany for an artificial intelligence initiative in the field of diabetes.

In a first step, Alcimed provided an overview on the major public and private data owners in the two in-scope countries. These data owners were then classified by our team in terms of data volume, quality, relevance and accessibility.

Based on these criteria, we prioritized some data owners as potential partners for our client and finally co-defined together a preliminary data acquisition strategy for this specific initiative.

One of our clients, a global medical devices player, wanted to test the market receptivity with regards to a solution they developed based on artificial intelligence and supporting the decision-making of anaesthesiologists regarding patients in coma.

Our team first assessed the strengths and weaknesses of the solution by setting up tests and by engaging with potential users in 9 Northern European countries. We also identified measures on how to overcome the identified barriers were equally identified.

After these steps, our team conducted a market sizing to assess the number of potential users and patients to benefit from this solution. Finally, we supported our client in understanding and navigating the purchasing process for such a solution.

One of our clients, a leader in the medical device industry, was searching for a competent partner in programming artificial intelligence applications for digital pathology, with the objective to extend its product offering.

To do so, our team conducted a global scouting of innovative AI providers, which were notably analysed in terms of technological maturity, business maturity and their network strength.

Based of our work, a selection of prioritized providers was approached to explore opportunities of collaboration with our client.

Our client, a public institution, wanted to support industrials players in its region through the use of AI in their production activities.

To do so, our team was called on to identify the potential usages of AI in production and to decipher how they could answer some key challenges faced by industrials.

Alcimed provided then recommendations regarding the modalities of AI integration in industrials processes (strategy to implement, stakeholders to involve, partnerships to set up, etc.) so that the potential usages identified and selected could quickly be operational.

A biotechnology stakeholder developed a new solution for diagnostics in oncology. They identified the analysis of the results as a remaining bottleneck in the diagnostic process, and wanted to offer a solution at this step of the value chain.

To help our clients, we developed a dashboard, that would convert simple questions into actionable insights. With our client’s input, as well as the insights of potential future users, we developed a dashboard that used state-of-the-art natural language processing (NLP), to create meaningful graphs and extracted statistical parameters from the complicated underlying data.

Our project enabled our client to get an insightful dashboard while guaranteeing data protection and avoiding hallucination from the generative AI model.

Our team supported an international pharmaceutical stakeholder in selecting a new AI tool to improve their knowledge management. Documents and data being stored on various platform, information retrieval was identified as a major hurdle in their activities.

To help our client, we first conducted internal interviews to identify the main pain points and challenges associated with knowledge management. Existing solutions were then mapped, characterized and prioritized.

Based on the investigation, we suggested an approach that would solve their current challenge, by leveraging previous unused functions of their current platforms, and with the integration of a new AI solution supplier.

Our client, a financial service provider, wanted to investigate the potential of AI models for the prediction of risk on their projects.

The project started by a mapping of the data available for predictions, followed by an extract, transform, and load (ETL) process for model building. Multiple models were then assessed in an iterative process. A final model was selected based on its performance.

The model was then adapted to be put in production and adapted to the client’s infrastructure and could then be used for decision making on new projects.