Artificial intelligence (AI) has been headlining the news lately – highlighting both the possibilities and the pitfalls. The global AI for life sciences market is booming, growing at a CAGR of 29.3% from 2022-20301. What will AI breakthroughs mean for pharmaceutical companies? We asked our first client dedicated to AI, Taylor Chartier, CEO/Founder of Modicus Prime, to share her perspectives.
What does Modicus Prime do?
Simply put, we enable computer vision for pharmaceutical quality control. Our proprietary software, mpVision, helps scientists independently train their own AI to achieve real-time AI classification of any imaging data – from biologic morphology analysis to commercial contamination detection – to ensure comprehensive quality control across both R&D and manufacturing.
What are today’s unmet pharmaceutical needs around AI?
Biotech, biopharma, and related initiatives in Big Pharma are looking for real-time quality assurance with faster data analysis. They also need comprehensive records, traceability, and explainability to ensure FDA compliance. Clients want high throughput processes in production for faster market results, as well as reduced operating costs with continuous processing. Finally, pharma companies want to maximize resources by eliminating manual, error-prone data curation and calibration in the lab and during production. These unmet needs highlight the necessity of self-service AI platforms that are fully in-house for pharma companies.
Can you provide an example of how Modicus Prime AI helps clients?
Pharmaceutical companies typically describe a specific issue in product quality that is causing significant cost and risk to drugs they are creating for patient use. These problems are extremely impactful operationally. In most cases, a company will have biologics susceptible to contamination or undesirable behavior during drug manufacturing. We show them how AI technology can help monitor the quality of the biologic and how they can train their own proprietary AI models to meet their unique quality assurance needs. Our mpVision platform provides the engine and tools to solve these issues using computer vision models that clients can train for use in production.
What are the benefits tied to quality around AI for life sciences?
Today’s pharmaceutical companies face a challenging therapeutic market with pressures around drug discovery and personalized medicine, plus risk management in a highly regulated industry. Regulatory bodies require Big Pharma to comprehensively understand drug product content during every stage of production. AI can directly address these regulatory stipulations by helping to lower quality risks while speeding throughput and reducing costs. To further maximize these benefits from AI, ISPE recently published AI guidelines in its GAMP 5 second edition, enabling pharma experts to ensure they that have validated AI solutions.
Computer vision is actually a mature AI solution in the medical device space, with over 250 FDA-approved medical devices already using computer vision technologies.
What about the risks associated with AI?
Life sciences risks around AI are often related to model biases. An AI model may not generalize well in the real world when a specific data set is biased towards a particular use case, or not truly representative of a statistical population distribution. Fortunately, such risks can be mitigated with a proper validation platform. The GAMP 5 second edition addresses these situations to help clients curate their data sets and properly validate their AI models. Our GxP-compliant platform, mpVision, provides an end-to-end validation tool to help clients achieve a representative data set with generalizable AI models. It is important to identify and implement the right technology and then adhere to industry AI guidance to mitigate quality risks.
Can you identify the next AI opportunities for faster, safer, more affordable drug production?
On the research side, we see tremendous development in helping drug researchers work more efficiently and consistently. Our AI technology also delivers benefits that improve the quality of their research. Traditionally, for every week scientists spend implementing software with imaging technology, they spend half of a week recalibrating imaging software, curating data and reporting. This process has been very time-consuming and prone to errors – approximately 50% of life sciences lab experiments are irreproducible and/or poor quality, costing $28 billion annually. With AI technology using generalizable image models, researchers can run models while saving time and effort – reallocating their focus onto experiments and away from trivial tasks.
Labs of the future will follow Pharma 4.0 to harness AI as an automation tool on both the research and commercial sides.
For manufacturing/production, clients want to replace offline, manual quality tests that have required running assays to ensure the efficacy, purity, and overall quality of the product. Now, AI methods can be used in-line during drug production to ensure product quality and reduce costs. Currently, if contamination is discovered in a batch, the FDA requires that the entire batch be discarded. Using an in-line monitoring process, the specific content can be detected in real-time, and operators can isolate the contamination, then release the non-contaminated product. This advancement helps manage risk and reduce costs.
The pharmaceutical industry is catching the wave of using in-line, real-time AI technology to achieve efficient, cost-effective commercial monitoring of the manufacturing process.
What questions do you commonly hear from life sciences companies?
We are frequently asked, “How can we as a life science company implement AI? Where do we start? What processes should we adopt?” Advanced algorithms carry incredible benefits but also carry huge responsibilities. Having explainable AI practices and a transparent process is at the forefront of our philosophy. mpVision is a GxP-compliant platform that constantly provides metrics to help users monitor and understand AI performance.
A life sciences company should be able to describe to auditors how algorithms were selected as well as the intricacies around the AI models. If you don’t have the capability in-house, employ a trusted AI vendor. Make sure you have addressed GxP compliance issues, and work with a system validation vendor such as Sware who is trustworthy – delivering services that are dependable for both research and production.
Founding Modicus Prime
“Ours was a customer-driven launch,” says CEO/Founder Taylor Chartier, “founded upon the belief that health is an inalienable right for all. Algorithm development took place during the COVID lockdown, with a small team delivering high-performance AI tools and training internationally. We succeeded in exceeding customer expectations while achieving a patented technology.”
How did Modicus Prime begin working with Sware?
We were describing our compliance initiatives to the private equity firm, Insight Partners. As a trusted service provider to pharmaceutical companies, we needed an automated quality management system to ensure mpVision’s GxP compliance. Insight referred us to Sware as one of its portfolio companies focused on automating validation pathways. This introduction helped us build a strong relationship right from the start.
How do you use Sware’s Res_Q Platform?
Res_Q has been instrumental in helping us become GxP compliant and structuring all our regulatory requirements and standards in a single ecosystem. This strategic implementation has put us ahead of other early-stage AI companies that don’t have the bandwidth to handle quality management intricacies. We appreciate that the Res_Q platform is seamless, so we don’t have to focus on managing compliance requirements and international standards.
Sware Res_Q technology frees us to be agile, providing GxP compliance to multiple customers for superb scalability – vital to a start-up company.
To discover how Sware’s Res_Q platform can boost your agility around AI innovation, request a demo today.
