Google Gemini, AI Commercialization in Pharma and More: Q&A with Abid Rahman, SVP, Innovation, EVERSANA

Abid Rahman
SVP of innovation
EVERSANA

Google recently released an updated version of its AI LLM. Gemini 2.5 Pro boasts a variety of new features that could help integrate AI even more into various workflows. It is still a tool, however, and in order to get the best results, it must be used properly. EVERSANA’s SVP of innovation Abid Rahman spoke with Pharmaceutical Executive about this updated LLM and how it can be incorporated into certain aspects of pharma.

Pharmaceutical Executive: What’s the Gemini 2.5 model from Google all about and what makes it appealing to innovators?
Abid Rahman: Gemini 2.5 Pro is Google’s most intelligent large language model (LLM) to date. Its release has generated significant interest across various industries and technology circles. Gemini 2.5 Pro is designed to enhance reasoning capabilities, enabling more complex problem-solving and informed decision-making. In most benchmarks, it currently leads in various tasks such as math, coding, and reasoning. It also has multimodal understanding so it can interpret and generate responses in text, images, and code.

What makes Gemini 2.5 Pro particularly appealing to innovators is its ability to process tasks step-by-step, leading to more accurate and context-aware responses. This feature is beneficial for complex tasks such as coding, mathematical problem-solving, and scientific research, where nuanced understanding is crucial. Additionally, this model has the largest context window of any commercial LLM so far with a 1 million token size, which allows it to consume thousands of pages of data. Its planned expansion to a 2 million token context window will enable the handling of more extensive data, further enhancing its utility in data-intensive fields.

PE: Are you aware of it being used across the pharmaceutical or life sciences industry today?
Rahman: It just launched, but I do know teams including ours at EVERSANA that are exploring various potential use cases of Gemini 2.5 Pro. For instance, it is being used for generative content creation, personalization, and operational efficiencies to help review and provide a deeper analysis on some clinical reports than previous models. As these highly capable models get included in other Google products such as NotebookLM and AI co-scientist, we will see them used in new drug discovery and scientific research. It’s early, but I’m confident we’ll continue to see use cases across pharma, especially as it helps to improve service delivery and provide better outcomes for clients.

PE: How is AI being integrated into commercialization strategies in Pharma?
Rahman: AI is being integrated into commercialization strategies across the entire pharma sector in many ways. From analyzing large datasets, generating insights, and automating processes, marketing content generation, content approval, there are so many opportunities in this industry to be transformed with Gen AI.

We like to say we’re helping to “pharmatize AI” for the industry, and there are more than 50 examples across our organization alone where AI is really transforming how we work through custom AI agents specifically built for pharma, to chatbots, AI-generated personas for training, image and content creation and more, the sky is the limit.

AI tools also provide sales representatives and MSLs with real-time insights and recommendations, improving interactions with HCPs. By analyzing data on HCP preferences and behaviors, AI assists in tailoring communication strategies, leading to more effective engagements.

All that said, there still must be a human component to ensure accuracy and that the AI-driven efforts are relatable and authentic, but it’s already proven we can use it for more personalized marketing content, optimizing sales strategies, and improving customer engagement. Additionally, AI can help streamline regulatory reviews, personalize content, develop hyper-informed treatment pricing models and so much more.

PE: What are the benefits of using AI in commercialization?
Rahman: Increased efficiency and productivity, improved accuracy in data analysis, and enhanced personalization of content are a few that come to mind. Think of it this way – as you bring a drug to market, the audience for the product can vary greatly by age, ethnicity, and language. Before AI was where it is today, custom campaigns were needed to reach people in different languages and cultures, and then custom campaigns needed to be deployed to connect with them. AI-powered analytics process vast amounts of real-world data, enabling pharmaceutical companies to identify effective market entry strategies. Machine learning models analyze prescribing patterns, payer coverage trends, and competitor movements to refine pricing and reimbursement approaches.

Several pharmaceutical companies have successfully integrated AI into their commercialization strategies. For instance, AI-driven pricing strategies have supported companies in optimizing pricing and improving competitiveness, enabling customized pricing approaches.

Today, AI can exponentially help in so many ways from image and content creation in seconds and suggestions on who to target based on personas. It’s powerful, but before it’s ultimately deployed, smart marketers must review to ensure it hits on the tone and experience the target audience expects.

PE: Are there any inherent risks of using AI in these situations?
Rahman: Without a doubt, the inherent risk of using AI in commercialization does exist. They include data quality issues, regulatory compliance challenges, and cultural resistance. Plus, there can be risk of bias in decision-making based on AI algorithms, as well as the potential for malicious use of AI technologies.

However, it's important to know that AI is here to stay, and we can’t ignore its impact and usefulness. Many of the concerns can be overcome by appropriate training and having the right frameworks and guardrails in place.

PE: From your perspective, do you see most pharma companies planning to build their own AI models, or do they hope to be able to use commercially available AI models, like Google Gemini?
Rahman: Pharmaceutical companies are increasingly integrating AI into their operations. The approach to AI adoption varies by company size and resources, and the type of AI solution needed. Most pharma companies are likely to use a combination of both approaches depending on the use case.

While some may build their own AI models to address specific needs and gain a competitive advantage, many will also leverage commercially available AI models like Google Gemini, CoPilot or ChatGPT to benefit from their advanced capabilities and scalability. This hybrid approach allows companies to optimize their AI strategies and achieve better outcomes in drug development and commercialization. There is also a distinction between a Generative AI model like Gemini 2.5 Pro and more custom-built machine learning AI models. Commercial GenAI models and LLMs can often be fine-tuned for business needs but most of the time they can be used with the right prompting techniques.

There is so much to learn about AI that we’ve heard often from clients “where do you start.” It’s best to have a trusted partner that can help navigate the AI highway.