For years, enterprise automation followed a simple formula. Find a repetitive task, automate it, count the hours saved and move on. In pharma, that approach left the hardest problems untouched. Data still sits in four different places, approvals still wait in someone’s mailbox, critical knowledge about what’s complete, what’s missing and what comes next still lives in spreadsheets, chat threads and people’s memories.

That model is no longer sustainable. Pharma companies are under pressure to bring therapies to market faster, operate leaner and stay ahead of regulators all at once. Every workflow must carry not just data, but the audit trail and decision logic behind it. Automating a broken process only makes it fail faster.

Featured Speakers

Bharadwaj (Brad) Popuri, Head of R&D Hyperautomation at Takeda, brings 14 years of experience in clinical trials, operational management and R&D operations, along with a research background in machine learning and cancer detection. He explains which pharma processes are ready to be rebuilt from the ground up and what makes agents context-aware in a GxP environment. He also covers why judgment must stay with humans.

Abhinav Kolhe, Vice President – Sales, Business Hyper-productivity at Persistent Systems, brings the view from across enterprise automation programs. He shares how to identify the right first use cases when opportunities seem endless. He also explains why AI delivers its real value inside a governed workflow rather than as a chatbot on the side.

What it takes is more than a better model

Models are increasingly interchangeable. What isn’t, is the work around them: redesigning the process, structuring the context and building controls, approvals and audit trails into the workflow. AI reads, summarizes and drafts. The process determines when an agent can act and when a human must decide.

Six Requirements for Reimagining Pharma Workflows:

  1. Question the process before automating it: With capital costlier and timelines tighter, the starting point is whether a process is designed for AI at all, not whether a task can be automated.
  2. Carry the audit trail with the data: Pharma systems must move the decision-making framework alongside the data, keeping records inspection-ready at all times.
  3. Choose processes that are truly ready: Connected systems, LLMs that read unstructured documents and humans in the loop now make it possible to rebuild multi-party processes, such as study startup, from the ground up.
  4. Fix data and process together: Fixing data alone delays outcomes, while fixing process alone leaves the foundation broken. An integrated, iterative approach delivers ROI within a year.
  5. Build context agents can trust: Agents become useful when they understand entity, state, history and the identity of the person asking, all of which are essential in a GxP setting.
  6. Keep humans accountable: Manual document handling and status chasing will fade as agents become proactive, but judgment calls and accountability will stay with people.

Join the conversation. Contact us at podcasts@persistent.com.

Speakers 

Bharadwaj (Brad) Popuri, Head of R&D Hyperautomation at Takeda

Abhinav Kolhe, Vice President – Sales, Business Hyper-productivity at Persistent Systems

View the Transcript