Why decision-making — not paperwork and process— determines speed, cost, and patient impact on drug development

While most AI adoption in biopharma has focused on drug discovery and preclinical research, roughly 70 percent of time and cost in drug development occur in the clinical stage — where productivity and regulatory success remain weakest.  Tanja Obradovic, as clinical leader at and now an advisor to HopeAI, argues that the industry has been looking in the wrong place for efficiency.  

"Preclinical drug development already strongly embraced AI as well as the area of post approval monitoring of the drug real-world performance.  The clinical development in the middle is the bottleneck where significant challenges remain with huge opportunity for efficiency gains in time and cost." — Dr. Tanja Obradovic, Oncology Medical Strategy Advisor of HopeAI

From trial-and-error to data-driven design 

According to Obradovic, the true value of AI lies in transforming decision-making in speed, availability of relevant documents, comprehensive evaluation of the future target product profile, how endpoints are selected, trials are design, and how probability of success is predicted.

“Nothing is as expensive as a failed drug, when you design well, everything goes faster and costs less.”  

A failed Phase 3 is a disaster — it reflects misunderstanding of earlier data and relevant external parameters/landscape, unsuitable approach to late development and that’s exactly where AI can make the biggest difference.”  

By combining historical data with modern algorithms, AI systems can evaluate hundreds of design and outcome scenarios that no human team could manually compare. This capability enables developers to cut years from development timelines while increasing the probability of success.

Learning from the past, at scale

“I’ve been to the FDA numerous times,” she notes. “Ten years ago, AI capability wasn’t there. Now it is and regulators are opening doors for much more sophisticated medical informatics AI use — and we must use it.”  

For Obradovic, the mission is deeply personal. She helped guide the clinical development of Keytruda, a drug that has treated millions of patients globally. Now she wants to ensure that next-generation therapies — from cell and gene treatments, small molecules to ADCs — reach patients faster through better design.  

Decision-grade intelligence, not automation 

HopeAI’s technology focuses on the core decisions that shape success or failure: 

• Clinical Development Plans (CDPs) and Pipeline: AI-supported frameworks that define endpoints, cross studies, rationale, and best/likely/worst-case scenarios. 
• Adaptive Design Simulation: Templates that compare multiple design options and operational assumptions across countries, sites, and enrollment speeds from early to late-stage development. 
• Probability-of-Success (PTRS) Modeling: Machine-learning models that estimate the likelihood a trial will finish on time — and achieve positive results — based on indication, phase, and design.

Changing the culture of clinical decision-making  

In most pharma settings, Obradovic notes, during development, or in-licensing/collaborations still rely on an expert's ranking projects by belief in limited available data. Even when marketing analysts contribute statistics, the process is “primitive as it lies upon market research that still is not using full power of AI.”   

"AI’s role in decision-making within an oncology clinical pipeline is very limited — and that must change if we want to cut timelines from ten to fifteen years as it stands now to at least half or more. When trials are aligned with patients and physicians, everything accelerates, and costs drop. Using large scale data and advanced algorithms to distill most critical needs in the real time during clinical development is a huge recent advancement." — Dr. Tanja Obradovic, Oncology Medical Strategy Advisor of HopeAI

From data to patients  

HopeAI’s mission aligns closely with Obradovic’s philosophy: to make clinical development a learning system that continuously improves with each dataset. By integrating clinical trials, real-world and literature-based evidence into intelligent design tools, HopeAI aims to help biopharma teams learn from every past trial — not to repeat its mistakes. 

"HopeAI helps new medicines reach patients faster; by using modern AI to learn from vast medical data and past trials, we can design smarter clinical studies and avoid costly mistakes. That means months to years saved — and lives saved in the real world." — Dr. Tanja Obradovic, Oncology Medical Strategy Advisor of HopeAI

About HopeAI

HopeAI is to bring hope to patients by accelerating the clinical development of new treatments through the power of AI. By integrating comprehensive clinical evidence with cutting-edge statistical innovation, HopeAI facilitates optimized clinical trial design, improves patient recruitment, enhances real-world, evidence-based decision-making, synthesizes data-driven support, and increases the probability of success in clinical development. For more information about HopeAI and its AI-driven solutions, visit hopeai.co

Media Contact

PR Team: [email protected]