Traditional and generative AI: an insight into the new normal for pharma firms

David Staunton

Head of Transformation, Life Science Manufacturing, Cognizant

Jacqueline Hora

Digital and Data Analytics Consultant, Life Sciences Manufacturing, Cognizant

Artificial Intelligence (AI) is transforming the way medicines are made. Expert knowledge and good data analysis remain vital to the life sciences equation, however.

When David Staunton, Head of Transformation, Life Sciences Manufacturing at Cognizant, is asked if AI will transform the way that the life sciences industry will operate in the future, his answer is: “It’s already doing it.” For example, AI — which excels at the super-fast processing of vast amounts of data — is already having a big impact on the way that medicines are discovered, launched and made.

Revolutionary impact of AI in pharma manufacturing

“Pharma companies have been able to synthesise around 1,000 molecules a year for drug discovery purposes,” says Staunton. “With AI, they can analyse 50 billion molecules a year. For context: if you count to 1,000, it takes about 15 minutes. If you counted non-stop to 50 billion, it would take you approximately 1,500 years. So, AI is completely game-changing for pharma companies.”

Every area of the pharma industry is feeling AI’s effects — including early discovery, clinical development, manufacturing, supply chain, commercialisation and even the digitisation of factories. Ultimately, that’s great news for patients who continue to benefit from new and more personalised treatments.

“AI significantly reduces drug development and launch timelines,” says Staunton. “As a result, new therapies will get to market faster.” AI is also altering the competitive landscape, with tech firms expected to increasingly enter the space traditionally populated by pharma companies.

Data-driven approaches to drug development

Staunton is, however, keen to flag up how ‘traditional’ AI (which recognises patterns in data) and ‘generative’ AI (which generates new content based on data) are used in the life sciences sphere. “There’s no doubt that generative AI can be super-helpful to pharma companies,” he says.

“However, the technology is not quite there yet. For example, it can’t make GMP (good manufacturing practice) decisions, and it’s not writing GMP reports without human-approved validation.”

Going forward, pharma companies will use AI to accelerate the development, manufacture and marketing of their products, says Jacqueline Hora, Digital and Data Analytics Consultant, Life Sciences Manufacturing at Cognizant. Developing a cohesive data strategy — which AI solutions consume — to generate insights is key.

AI isn’t a replacement for life sciences knowledge.

Well-crafted data strategy can help manufacturers

Hora points out that the more meaningful and contextualised the source information that an AI solution has access to, the better insights it will generate.

“Contextualising your data is critical to understanding past performance and using historical data to predict future outcomes,” she says. “Contextualising data starts when equipment and process systems are configured, and it typically involves combining details from multiple sources to build a comprehensive picture of events.

“A well-crafted data strategy will enable this seamlessly, ensuring data from your systems are easily accessible and usable. In AI solutions, the foundation for generating accurate and dependable information and insights lies in the availability of high-quality and reliable source data, instilling a high level of confidence in their accuracy and authenticity.”

Value of quality data and expert interpretation for AI purposes

Cognizant is currently working on a use case where generative AI will form part of a solution to provide advisory support to new technicians. Key to the success of this solution is the capture of previous knowledge and experience of manufacturing operators to use as a knowledge base.

“There is an impression that AI can fix things when it can’t,” Hora states. “If you put bad information into your system, you’ll get bad information out.”

AI isn’t a replacement for life sciences knowledge, says Staunton. Instead, expert data analysis is — and will remain — crucial. “You’ll have to be able to read and interpret the data AI gives you in order to understand if it’s sound or not; this will become a key people skill,” he says. “Our increasing use of AI means that data is only going to become more important in the years ahead.”