In my earlier blog, I highlighted a few considerations that are critical for the adoption of Generative AI, aka ChatGPT in Life Sciences. One way to look at these considerations is to view them as constraints that need to be overcome.
In this blog, I want to cover key use cases that by their nature lend themselves more easily to the usage of Generative AI. This blog needs to be read keeping the broader business processes in mind.
Let me start with two specific use cases where we are already seeing active adoption of Generative AI in Life Sciences:
In the experience economy, pharmaceutical companies face the challenge of building personalized customer journey maps in an omnichannel world. The industry has been searching for an efficient approach to address this. One approach that the industry has taken is to build a decision engine (Next Best Action) using traditional AI/ML models. With Large Language Models (LLM), there is now a new way to tackle this challenge. ChatGPT is effective in creating personalized journey maps and associated omnichannel plans for a defined customer segment. This saves a significant amount of time and effort while offering a cost-effective option vis-à-vis traditional methods that require high investment.
By carefully setting up the context and guardrails, you can ensure that the technology understands the intent of the question and provides consistent responses. With this foundation in place, ChatGPT can then be used to develop detailed journey maps and plans that cater to each customer segment and persona.
Conversational AI / Chatbots have been around for a long time. Historical benchmarks on the quality of conversations, the cost, and implementation time have hindered more widespread adoption. This is not unique to life sciences alone. Even virtual assistant technology applications like Siri, Cortana, Alexa, and a multitude of chatbot companies have been successful but to a limited extent. Now, Generative AI offers you the ability to build your own chatbot and scale it to the next level. One example I was personally involved in was building a robust interactive chat functionality in a matter of days instead of months.
ChatGPT can serve as a channel for smarter pharmacovigilance (intake of adverse events). ChatGPT acted as a chatbot to capture crucial adverse event (AE) information from patients (including follow-up for missing pieces of key information). ChatGPT also converted unstructured conversation into specific data elements in the required format so that downstream systems (Oracle Argus Safety, for example) can ingest this data, allowing faster case processing and reporting of adverse events.
ChatGPT can double up as a call centre/chat agent in other domain areas, with a carefully defined context preventing the AI from providing incorrect information. By guiding the conversation and asking relevant questions, ChatGPT can collect valuable information from users (patients) in real-time.
The true power of ChatGPT lies in its ability to parse sentences and extract the required information from users' responses, even if they are not formatted as data inputs. Once the AI engine has collected all the necessary information, it can be converted into a structured format that can be easily fed into a database or system.
The above two use cases are just the tip of the iceberg. We are actively experimenting with ChatGPT for a few other use cases, and are seeing promising results. While some of these are in the proof of concept (POC) stage, we see many maturing fast.
The practical applicability of Generative AI is almost endless, and given the enormous potential of Generative AI, the life sciences industry is poised for an exciting journey ahead. Using ChatGPT prudently, you can accelerate your organization’s digital journey and significantly improve future-readiness as new technologies catch up (and probably surpass) ChatGPT.
But, for life sciences companies, the key is to go the distance. To use the fishing analogy in this context, it’s more important to learn how to fish than just get the fish; so, it’s more important for life sciences leaders to learn how to leverage Generative AI to its full potential than just buying or taking a piecemeal approach to using it.
In my next blog, I will focus on how to quickly start implementing ChatGPT and scaling the projects to sustain success.