At a Glance
The rapid growth of AI Is a game changer in practically every aspect of life. And there is no doubt the biotechnology and pharmaceutical industries can benefit hugely from it. Our teams have identified three key challenges that need to be addressed to ensure success:
- A lack of understanding of AI.
- Data constraints.
- Understanding the talent you need.
The article also details how best to formulate a strategy and approach to ensure you successfully integrate AI into your organization.
AI offers biotech and pharmaceutical companies countless opportunities
Novel algorithms, greater computing power and huge datasets will support dozens of applications across the value chain. Pharmaceutical R&D can benefit too, in areas such as:
- De novo small-molecule design.
- New biomarker identification.
- Targeted therapy development based on genomic data.
- Drug interaction predictions.
- Permeability and patient response.
Commercial and medical units in pharmaceutical corporations can leverage new computational techniques with:
- Comprehensive KOL mapping.
- Landscaping algorithms.
- Virtual patient engagement assistants.
- Marketing predictive analytics.
- Sales force optimization.
- Nonlinear relationship estimation with HEOR real-world-data.
But if you are to successfully tap the true potential of AI then we need to overcome the following three key challenges.
1. Decision makers lack an understanding of AI’s capabilities and limitations
This is understandable. It’s a new and still immature technology. And there are no established conceptual frameworks or strategies to help understand where the value lies. Or, for that matter, how it could fit into an organization.
Remember, AI is not a new form of IT.
In our experience, this leads to failure. That’s because IT can be summarized as automating information logistics (storing, sorting and transporting). However, AI automates much more complex cognitive tasks. For example, it’s the difference between an airplane’s autopilot system and the cockpit dashboard. The former is AI, as it is making decisions about direction, speed and altitude. Whereas the dashboard is presenting information and data – that’s IT.
2. Organizations face significant data constraints
Most AI solutions depend on historical data. And that can often be lacking, or of poor quality. Equally, it may not be available due to legal hurdles. Acquiring quality data requires long-term, sustained and cross-functional effort, and that’s where many organizations struggle. To scale capabilities you need to implement and adhere to data capture points across workflows. And that means your people have to care about this and put the effort in to ensure it happens
The right data may not be available internally.
You may need to consider strategic partnerships, acquisitions or licensing agreements to build up the right data assets. This needs to become a priority for senior management if the organization wants to unlock AI’s true potential.
3. Understanding the talent and structure needed to implement AI
One common misconception is the belief in an ‘AI expert’. This all too often leads to a costly arms race to equip your organization with the one person who can unlock AI’s value. More often than not this leads to disappointment.
It’s not about finding one person with specific technical skills.
It’s about having a team that can navigate the big picture, based on a thorough understanding of technical capabilities. Many organizations already have what is now called ‘AI’ in-house. They just don’t realize it. Because the underlying enabling technologies are not new. So rather than build new technical silos, empower existing teams to build the foundations that can accommodate AI and integrate it with their industry expertise and corporate vision.
AI demands a strategy, not a one-off initiative
Considering the three limiting factors together will help you create a coherent approach to AI. Remember, reaping AI’s benefits is a long-term effort at the intersection of fast-moving, converging technologies. This calls for a strategy, to manage risks, limit costs and navigate through uncertainty towards real value.
Start with opportunity discovery
This requires analyzing the AI landscape, considering your data resources and technology against organizational requirements. By doing this your team has the tools to brainstorm, evaluate and map opportunities into three categories:
- Opportunities to implement immediately
- Options for building data, technical and or talent foundations
- Ideas that will benefit from strategic pathfinding.
Avoid focusing on quick-wins or lighthouse projects. These are often chosen for feasibility and distract from the long-term sources of competitive advantage.
There are four phases to implementing immediate opportunities, Proof of Value (PoV), Proof of Concept (PoC), implementation and, once operational, life cycle management. Let’s discuss the first two stages in more detail.
1. Proof-of-Value (PoV)
A quick but vital exercise where you explore plausible corridors of technical performance. Question the potential value to the business? What will it depend on? Remember, it’s all too easy to get carried away by new technological possibilities. So it’s essential to look at ROI. Doing PoV allows the Proof of Concept to be conducted in a focused manner. This is because key value drivers often only become clear when looking at ROI in a structured way.
2. Proof of Concept (PoC).
This is the most critical step. It brings science and its methods to data science. PoV establishes the capacity of AI tools to deliver value within the scope of the opportunity. The ideal way of executing the PoC is via sprints that each finish with go/no-go decision gates. This enables rapid re-categorization of ideas that lack feasibility, mitigating risks and costs. If an opportunity cannot be implemented immediately during PoV or PoC, it could be because your organization lacks foundational capabilities to execute it, or AI cannot deliver on the expected use case.
Establish the right foundations and create data readiness
Achieving this requires the orchestration of various lines of work:
- Historical data needs to be excavated, gathered and curated to assure quality.
- Adjust IT systems to facilitate information capture.
- Remodel team workflows to prioritize data capture.
- Train people to understand data science and its value.
Finally, build a communication and scoping bridge between AI specialists and the rest of the organization.
What if your technology cannot deliver the desired results?
Be disciplined in strategic pathfinding to keep the aim on the strategic horizon. Your biggest opportunities may lie in technologies that are just emerging. Strategic pathfinding integrates AI technology monitoring and evaluation with a roadmap to build enabling and foundational requirements. Through pathfinding, you can prepare your organization to be in pole position for AI applications that blossom in three, five or ten years.
Ultimately, be aware of challenges and have a plan to overcome them
The range of AI use cases in the pharmaceutical and biotechnology industry is rapidly expanding. From the application of old-school decision trees, drug adherence prediction and even learning models that support or reject biomedical hypotheses; opportunities abound. However, to successfully leverage opportunities, make sure you are aware of the challenges and have a plan to overcome them.
Interested in any aspects of AI? Like further information on how you could integrate it within your organization? Then, please get in touch with our team today.
About the authors
- Dr. Paul von Bünau, Managing Director, idalab
- Sara Enriquez, Client Service Partner and Principal, a-connect
