At the recent Big Data & AI World in Frankfurt, Midas Pharma shared a narrative of transformation that resonates with any organization buried under manual reporting.
For Midas, a global B2B leader in the pharmaceutical value chain, the stakes are high: they manage everything from active ingredients to manufacturer audits and regulatory submissions.
The Starting Point: Escaping “Excel Hell”
In the beginning, Midas Pharma faced a classic "Excel Hell."
As Max Maelike, Head of Data Science, recalled, the complexity of the B2B pharma world meant that every department had different requirements. Before ThoughtSpot, this led to a culture of ad hoc Excel analyses that were outdated as soon as they were saved.
The business risk was tangible. "Identifying new value-generating opportunities was mostly a matter of luck," Max noted. There was no single source of truth, and hours were wasted cleaning data rather than analyzing it.
The Evolution: From Reports to “Agentic” Insights
When Midas first integrated ThoughtSpot with their Snowflake Data Cloud, the goal was simple: speed and self-service. They achieved that almost instantly, with 99% of queries now returning results in under a second.
But the journey has since evolved from search-based analytics to agentic analytics.
Then: Users searched for data to build their own reports.
Now: AI agents show data and explain the reasoning behind it, providing insights.
"Agentic AI now enables users to get answers to advanced questions themselves. The system shows them why the complexity exists and how to approach the problem—sometimes even inspiring entirely new ways of thinking."
Sarah Büttner
Data Scientist at Midas Pharma
Trust as the New Currency
A major evolution in Midas’ strategy has been the focus on AI transparency. In the early stages, "hallucinations" were a major concern for the data team. By utilizing ThoughtSpot’s ability to show the "search tokens" and the data foundation behind every visualization, Midas has built a high-trust environment.
"The ability to independently verify whether a result makes sense has created a strong level of trust. Potential hallucinations can usually be identified very quickly by users themselves."
Sarah Büttner
Data Scientist at Midas Pharma
Quantifying the Transformation
The shift from a lean IT-led model to a true self-service culture is best seen in the numbers:
| Metric | Impact & Evolution |
|---|---|
| User Empowerment | 50 Liveboards created by business users vs. only 10 by IT. |
| Productivity | Average time savings of 3–4 hours per week per user. |
| Speed to Value | 72% of initial users were active and self-serving within just 4 months. |
| Engineering Efficiency | Data Vault 2.0 and automation allow the team to stay lean while supporting 10,000+ monthly visualizations. |
These efficiency gains have changed user behavior. Because they no longer have to wait minutes for a Python script or an Excel macro to run, users are experimenting more.
“Previously, all energy was consumed by preparing and cleaning data. Now, our experts finally have the time to be the experts they were hired to be. People now experiment much more and are willing to try complex searches because they no longer have to wait."
Karl Matzenbacher
Data Engineering Lead, Midas Pharma
The Road Ahead: 2026 and Beyond
The original goal was to fix reporting; the new vision is a "road expansion." The Midas Pharma team is no longer just maintenance workers for an analytics backlog; they are the architects of a data-literate culture.
By the end of 2026, Midas Pharma intends for AI to be a "confidently and naturally" used tool for every employee, ensuring that no business opportunity is left to luck again. They haven't just defeated the Excel monster; they’ve built a data engine that moves as fast as the pharmaceutical industry demands.
See how you can do the same with your own data - schedule your ThoughtSpot demo today.
