A web-based dashboard and process design “toolkit” with “ChatGPT” like functionality that incorporates analysis of process data to determine critical parameters for CAR-T based therapy development, avoid bottlenecks, and ultimately connect patient clinical responses to key process parameters; The tool will allow visualization of the most impactful process parameters using each of the seven AI algorithms in a user-friendly interface. In addition, the interface will provide the option to perform simulations to further understand how changes to process parameters can potentially impact patient responses.

InsightsTM Explainable ai

AI/ML-Driven InsightsTM

The utilization of seven well-established AI/ML models allows for in-depth analysis of critical process parameters. This sophisticated analysis goes beyond human capabilities, enabling the identification of subtle patterns or correlations within the data that might be pivotal in understanding and optimizing cell therapy processes.

Connecting Clinical Responses

By establishing connections between patient clinical responses and key process parameters, the platform enables a deeper understanding of how treatments affect patients. This linkage can accelerate the refinement of therapies, potentially leading to more targeted and effective treatments for patients.

Interactive AI/ML-based Process Development Simulator

An AI-based process simulator for cell therapy would be a powerful software tool designed to model and simulate various aspects of the cell therapy manufacturing process, allowing researchers, scientists, and engineers to optimize and refine the production of cell-based therapies.

AI Explainabilty

For those tasked with the responsibility of providing explanations to regulators (e.g. the FDA), AI/ML remains a bit of a mysterious “black box” where inputs go in, outputs go out, and explaining what happens in the middle is a challenge. We make AI/ML readily explainable to those who are not experts in the field.

Efficient Data Utilization

By transitioning from disorganized data storage to a centralized, web-based app that leverages storage in a structured SQL data base, the platform will enable efficient utilization of vast amounts of data that were previously underutilized or not fully analyzed. This optimization of data analysis can reveal patterns, correlations, and insights that were previously inaccessible, potentially expediting the development and refinement of cell therapies.

Visualization and Decision Support

The web-based dashboard offers visual representations of analyzed data, providing an intuitive interface for understanding complex information. This aids in quick decision-making and facilitates the identification of bottlenecks or areas for improvement in the cell therapy development pipeline.

ChatGPT-Like Functionality

We incorporate “ChatGPT” like functionality to enhance the user experience by providing an interactive interface that interprets and presents complex data insights in a conversational manner. This not only simplifies the understanding of critical parameters but also bridges the gap between data analysis and actionable decision-making.

Leveraging Generative AI for Reliable Data Synthesis

In some cases, the available data from clinical trials may pose some limitations in terms of significant patient numbers for analysis. We use generative AI to augment small data sets to help build more reliable more reliable models. This approach uses the knowledge learned from larger datasets and adapts it to generate synthetic data in a specific domain.

Conversion and analysis of handwritten batch and medical records into digitized formats

Much work in the world of process development is recorded in the form of hand-written batch records that often remain in lab notebooks never to reviewed or seen again. As such, much information that could be garnered to enhance the process may be lost. We incorporate generative AI to extract and convert this hand-written information into a format that can be included in client models.

Omics Visualizer

Integration of various omics data types, such as genomics, transcriptomics, proteomics, and metabolomics.