Read the full article in RNZ
We’re thrilled to share that Rhein-Neckar-Zeitung (RNZ) recently interviewed our founders, Dr. Christian Aichmüller and Dr. Manasi A-Ratnaparkhe, in a deep dive into PAICON’s mission to transform cancer diagnostics and therapeutics through inclusive, AI-powered solutions.
In the interview, the founders discuss the inspiration behind PaiChat, our AI tool designed to offer second opinions for cancer diagnosis, similar in function to ChatGPT—but trained specifically on histopathology images. The foundation of PaiChat is PaiX, a diverse datalake built from over 3 million images across 50+ countries. Built on PaiX, PaiChat is designed to offer second opinions in cancer diagnosis by analyzing stained tumor tissue images—even in resource-limited settings.
The interview also covers:
- Why diversity in phenotypic and epigenetic data is critical
- The collaborative roots of PAICON, from DKFZ to EMBL
- Our award-winning trajectory, including the Heidelberg Startup Award and DMEA Nova Award
- How PaiChat is being developed to work reliably even with imperfect or low-resource imaging data
Driving Global Impact with Scientific Rigor and Local Insight
The RNZ feature also spotlighted how PAICON is bridging cutting-edge research with on-the-ground realities. With scientific roots at institutions like DKFZ and close collaborations with clinicians and partners in regions such as India and Cambodia, we ensure our solutions are not only technically advanced but also context-aware and globally relevant.
What sets PAICON apart is our commitment to data diversity, transparency, and ethical collaboration. Our data lake includes multimodal inputs from across geographies, and we work to harmonize formats, languages, and documentation standards because scalable AI begins with structured, inclusive data.
“Communities must trust that we’re building with them, not just for them.”
— Dr. Christian Aichmüller
Through this approach, PAICON isn’t just building products—we’re building an ecosystem where AI can truly serve all populations.
