Yseop--a software company specializing in artificial intelligence (AI) and natural language processing (NLP) technology—has announced a collaboration with Eli Lilly and Co. The partnership centers around Lilly’s deployment of Yseop’s Augmented Analyst automation platform, in order to help accelerate the time to market for novel therapies.
To learn more about the partnership and the technology at its center, Outsourcing-Pharma connected with Nouri Chibane, head of sales at Yseop.
OSP: Please tell us a bit about Yseop—who you are, what you do, key capabilities, and what sets you apart from other companies operating in this same sphere.
NC: Yseop (pronounced “easy-op”) is a world-leading AI software company and pioneer in NLP. Yseop’s expertise lies in data analysis, machine learning, and language technologies.
Its industry-leading Augmented Analyst platform analyzes enterprise data and delivers insight and document automation that supports Augmented Financial Analyst and Augmented Medical Writer applications. Our goal is to maximize efficiency by automating complex business processes and liberating people from extremely tedious and repetitive tasks.
Pharmaceutical companies like Eli Lilly rely on Yseop’s pioneering natural language generation (NLG) AI to scale human expertise through automated content generation. Generating clinical study reports (CSRs) is a very tedious step before submitting new drugs for approval, taking weeks and in some cases months to produce. Automating CSR writing with AI-based technologies, like Yseop, saves significant time, reducing report writing times by an average of 40%.
OSP: Please share your perspective on how the use of AI and other advanced analytical tech has evolved in drug development, particularly the approval process.
NC: Any advanced technology follows a similar life cycle starting as a labor-intensive kernel to implement, needing to be configured in a custom way for each application. Over time the technology is productized and matures to become more usable. This means a dramatic shift from only customizations done by engineers to little to no configuration that anyone can use. Yseop’s Augmented Analyst has followed that same trajectory.
Specifically, regarding the approval process, companies are using AI and analytics to reduce errors and shorten cycle times. For instance, a medical writer can become much more productive when paired with a bot to draft their regulatory submissions. The draft is generally error-proof and produced in under a second.
OSP: Specifically, what have been some of the notable milestones and developments, and some of the common stumbling blocks?
NC: The major milestone recently was the release of our no-code studio. Now the end-user, like the medical writer mentioned above, can directly interact with the AI to get the most out of it. It took years for us to develop the right methods and techniques to allow the effective use of AI by a non-technical person.
The classic challenge for our customers is data quality. If data is poorly organized and not well-structured then it’s going to be hard to process it. Most companies have done the hard work over the last several years to upgrade data systems and I’m happy to say that most are well positioned in terms of data quality.
Another major milestone is our recently announced strategic collaboration with Lilly, which allows them to deploy Yseop’s world-class enterprise automation platform, Augmented Analyst, to accelerate bringing Lilly’s medicines to patients. With the agreement, Lilly will leverage Yseop’s Augmented Analyst to transform data into high-quality narratives and regulatory submission reports, at scale and error-free.
In addition to quality improvements, Yseop empowers users to focus their time on more impactful initiatives. Together, both companies will further develop the Yseop automation platform to expedite the drug approval process to realize time and cost efficiencies for Lilly.
OSP: Please tell us about your Augmented Analyst platform—how it works, and what it can do to transform operations in various ways.
NC: Augmented Analyst is a platform that runs applications we call automation acks. Each automation pack performs a specific task like automating a document type. A large organization will want different types of automation tasks and the platform orchestrates the natural language technologies and other AI that we need to deliver for our customers.
An automation pack helps a user automate a report from data. To do that, you need data modeling and transformation, data analysis to decide what to write, ontologies to understand how to talk about the data, user permissions, and an interface to control it all. The automation pack brings that together.
OSP: Can you share any specifics about how Augmented Analyst might be put to work at Lilly?
NC: Eli Lilly is using our technology to automate reporting across the enterprise. One example is regulatory submission drafting, which saves time and reduces errors. Different functions in the org will use our writing bots to help employees be more productive while offloading tedious tasks. As an example, portions of the eCTD can be produced in seconds rather than weeks or months using our technology and that is very attractive to our customers.
OSP: What are some key questions a sponsor like Lilly or their research partner should ask before diving into a solution like your platform?
NC: Data and change management are two areas to think about from the beginning. Do I have relatively clean structured data? How can I prepare my organization to adopt automation most effectively?
Many organizations have existing processes, built years ago and based on different paradigms. Introducing automation should always be done with that in mind.
As an example, when writing errors are reduced through automation then you often need less QA. Rethinking document drafting QA cycles is one example of something a sponsor or a research firm can contemplate when they set out on this journey.
OSP: Do you have anything to add?
NC: Bots are in the workplace and that trend is only increasing. People tend to be creative thinkers when compared to a machine but when it comes to crunching numbers, a machine does a better and faster job. Companies have been accumulating huge amounts of data and they’ve largely asked people to manage and interact with it. Sometimes it is unpleasant and inefficient to ask someone to work with large data sets directly.
Technologies like Yseop help free people from tedious work while allowing them to accomplish more. Automated generated data reports help speed up the production of resources while reducing costs and risks of human error. Ultimately, this type of technology helps ensure accurate and high-quality reporting, while streamlining processes to produce reports at a faster rate, giving teams more time to focus on valuable and strategic tasks. The goal is to augment your team’s medical report writing expertise at scale.