340+ researchers. 7 international sites. A looming retirement wave threatening to erase 25 years of accumulated solar cell expertise. This is the story of how Wissenswerk helped SolarTech Global turn a crisis into a competitive advantage.
SolarTech Global is a multinational solar energy research organization headquartered in Munich, Germany, with facilities in Spain, Japan, the United States, Brazil, India, and Australia. Founded in 1998, they pioneer next-generation photovoltaic cell technologies — from perovskite tandem cells to degradation-resistant module coatings.
Their research teams span materials science, electrical engineering, module testing, and field deployment. Each site has developed its own deep specializations over two decades — and that's precisely where the problem began.
In early 2023, SolarTech's leadership realized they were facing three converging crises:
42 senior researchers — roughly 12% of the entire workforce — were scheduled to retire within 18 months. These were the people who held 25 years of institutional memory: why certain cell architectures were abandoned, which coating formulations failed in tropical climates, and which testing shortcuts actually worked.
Each of the 7 sites had developed its own documentation culture. Munich used Confluence, the Osaka lab kept handwritten lab notebooks scanned as PDFs, and the Austin team relied on shared Google Drives with 8,000+ files in nested folders. Cross-site collaboration meant emailing attachments and hoping someone remembered where to find them.
The São Paulo team spent 4 months investigating a perovskite formulation that the Bangalore lab had already proven unstable 3 years earlier. The findings existed — buried in a PDF on a local server nobody outside India could access. This kind of duplicate work was costing SolarTech an estimated €2.4 million per year.
SolarTech's CTO, Dr. Anna Lehmann, championed Wissenswerk after a proof-of-concept with the Munich headquarters. Here's how the rollout unfolded:
Wissenswerk's team conducted a knowledge audit across all 7 sites. They identified 34,000+ documents, 2,100 scanned lab notebooks, and 6 different documentation platforms. A priority matrix was created, ranking knowledge assets by risk of loss and strategic value.
Wissenswerk integrations connected Confluence, Google Drive, and SharePoint instances. Scanned lab notebooks from Osaka were processed through Wissenswerk's visual knowledge extraction — diagrams, handwritten annotations, and experimental formulas were all captured. By week 8, 28,000 documents were indexed and searchable.
The 42 soon-to-retire researchers participated in weekly AI debrief sessions. In 5-minute conversations, Wissenswerk's AI asked context-aware questions about their specializations — why certain experiments were stopped, which suppliers to avoid, and what undocumented tricks they'd learned. Over 12 weeks, 680 debrief sessions produced 4,200+ structured knowledge entries.
All 7 sites went live simultaneously. Researchers could now ask Astra natural-language questions like "What perovskite formulations have been tested for tropical humidity resistance?" and get instant answers with source citations — whether the knowledge originated from a Munich Confluence page, an Osaka lab notebook, or a verbal debrief with Dr. Tanaka.
Two moments crystallized Wissenswerk's value for the entire organization:
A junior researcher in Austin asked Astra about tandem cell adhesion under thermal stress. The answer cited a handwritten lab notebook from Dr. Kenji Yamamoto in Osaka — scanned 6 years ago and never read by anyone outside Japan. It contained a coating technique that solved a problem the Austin team had been struggling with for months. The fix was implemented in 3 days.
Dr. Yamamoto documented a silane coupling pre-treatment that improved adhesion at 85°C by 340%. The technique was tested on 120 samples across 6 weeks with zero delamination failures.
Wissenswerk's conflict detection flagged that a new module specification from the India team contradicted safety testing results from the Australian facility. The India spec called for a thinner encapsulant layer to reduce cost, but Australia had documented that anything below 0.45mm failed accelerated aging tests. Without this catch, an 18-month joint project with a hardware partner would have launched with a flawed design.
Within 8 months of full deployment, SolarTech Global measured the following outcomes:
"When Dr. Yamamoto retired in October, I thought we'd lose 22 years of coating expertise overnight. Instead, a junior researcher in Austin found his technique three weeks later through Astra — and it solved a problem we'd been stuck on for months. That's when I knew Wissenswerk had changed everything."
Wissenswerk