Humans × Agents: The Formula for Efficiency in Pharma 

In a recent episode, Nataliya Andreychyk spoke with Thomas Mrosk and Manuel Mitola, founders of ctcHealth, about integration of AI agents and their role in highly regulated pharma marketing workflows.  

Take Baby Steps. Then Scale 

“Look at the infrastructure you already have,” said Mrosk. You need to understand what is missing and whether your new agent will integrate with your existing ecosystem.  

Too often, companies begin with grand AI ambitions meant to overhaul entire workflows. But that’s too much stress that can impact your organizational health and result in low adoption rates.  

“Apply a fail-fast philosophy,” added Mitola. “Run small pilots. Test if these experiences bring value. Then scale what works.” It is not about replacing people. Your task is to find mundane tasks that agents can handle reliably. 

Once you see an agent demo, it’s tempting to think of replacing entire roles. Yet, it is important to remember that human work, especially in pharma marketing, is layered and nuanced.  

“If you want to build an agent for a brand manager,” said Mitola, “don’t try to make it the brand manager. Focus on tasks the brand manager performs, like preparing parts of the brand plan or summarizing performance insights.”  

Context Is Everything 

Pharma doesn’t need “just another AI tool,” said Nataliya. It needs agents that understand context, like brand templates, compliance rules, approved assets, and data housed in systems like the DAM or CRM. Since the agent acts in a regulated space, context helps them keep everything compliant and on-brand.  

For example, eVa, Viseven’s AI agent, ensures compliant production by generating content from pre-approved modules and, if configured, delivers output aligned with your brand’s voice and rules. It also checks the new materials against compliance rules before MLR and gives you recommendations to improve them.  

Even with largely accurate, consistent outputs from context-aware agents like eVa, humans are still essential to keep errors at bay. Technology helps break the silos between people involved in content creation.  

Examples of Agents in Action 

For medical reps, AI agents can transform how data is captured and used. After all, entering every interaction manually into a CRM is tedious and prone to errors.  

“What if reps could just dictate what happened on the call?” suggested Mrosk. “The agent could structure it, ensure compliance, and push it to CRM.” Once you feed the agent with your data, it can do much more, like identifying the biggest trends in a territory or suggesting where to focus next. That’s Next Best Action (NBA) done right. 

Launching a new brand isn’t a single project. It’s a collection of dozens of interconnected tasks. “At ctcHealth, we see agents as an ecosystem,” said Mitola. “Each agent tackles a sub-task, so marketers get consistency in a way it collects and uses data.” 

What is also exciting is how agents act as collectors of a company’s intelligence. They capture best practices from every affiliate and make it possible to replicate success globally. 

Where AI Assistants Still Struggle 

Even the smartest agents hit the wall when it comes to replacing human thinking and domain expertise. To make agents useful, you need to feed them with your knowledge base and craft comprehensive prompts. 

Yet, there are still the risks: hallucinations and compliance slips. Testing protocols, technical guardrails, and clear disclaimers are essential to reduce them.  

The guiding principle behind every agent is simple: pair human expertise with AI. While agents’ contextual understanding is essential, a human in the loop is key to making sure agents perform well in a regulated environment. 

The key takeaway is this: AI agents won’t replace people in pharma marketing. They exist to amplify how experts think, decide, and deliver value. 

---