Gombe AI – After decades of handwritten field notes and a growing digital backlog, the Jane Goodall Institute is using large language models and a new cloud tool with Amazon Web Services to digitize and translate more than 500,000 pages of chimpanzee research—aiming to m
On the ground in East Africa, chimpanzee research starts the way it has for decades: with careful watching and a clock. Primatologist Jane Goodall and a team of researchers have spent more than six decades quietly observing chimpanzees with binoculars in East Africa.
Field researchers take handwritten notes every 15 minutes when observing a single chimpanzee—and every minute when observing mothers and infants. Then comes the slower work: the notes must be digitized, in multiple languages including English and Swahili. Lilian Pintea. the vice president of conservation science at the Jane Goodall Institute. said it typically takes up to two days to manually enter all field data into web-based systems.
That labor has produced a multi-year backlog. The institute has operated with data awaiting upload since it created its first digital database in 1997.
In 2025, Pintea began using large language models to accelerate the digitization of more than 500,000 pages of handwritten notes. The goal is practical—making scientific data more searchable for researchers at the nonprofit Goodall founded in 1977. Pintea said preserving Goodall’s legacy by digitizing her handwritten notes has become a priority as well after her passing in October 2025.
“AI is just a continuation of our long history of different technology cycles,” Pintea said.
The institute’s challenge isn’t only how quickly notes can move from paper into a computer. It’s also where all that information lives. Pintea said one of the biggest challenges the Jane Goodall Institute faces is storing a large, disparate set of data spanning five generations of chimpanzees.
When university researchers visit Gombe. their findings—physical notes. photographs. and audio and visual files—are stored in their own institutions’ databases. The Jane Goodall Institute is building a new data tool with Amazon Web Services called the Gombe AI Research Platform. Pintea said the aim is to store the information in a single. more accessible place. making it easier for researchers to collaborate.
Digitizing at the scale the institute is working at has required manpower. Pintea said a team of nine principal investigators—each supported by two to three Ph.D students and roughly half a dozen undergraduate students—helped the institute digitize these assets.
The partnership with AWS also took a winding path. It began in March 2025 when Goodall delivered a keynote at the AWS Imagine Conference in Washington. During that event, Pintea said he was first asked to explore a potential collaboration between the nonprofit and AWS’s cloud computing platform.
In April 2025, during a kickoff meeting for the collaboration, the institute wanted to use AI to identify medicinal plant use in video footage and classify which plants the apes were eating. That use case was deemed too rare to pursue further.
By June 2025. the two organizations had pivoted toward a more widely helpful idea for AWS’s AI: photo and video record analysis. Taimur Rashid. the managing director of the Generative AI Innovation Center at AWS. said AWS and the institute decided to build on an existing video search tool called WISE. developed by the University of Oxford in January 2023.
Rashid said AWS used WISE’s conceptual framework to develop a cloud-based tool for visual search and computer vision models to analyze chimpanzee footage. so researchers wouldn’t have to spend hours manually screening videos. Rashid described the bigger promise of turning long-standing observations into searchable knowledge: “We can take all of these years of analog data. digitize it. make it searchable. and add elements of gen AI to look at these data points and really comprehend all these years of data.”.
By August 2025, an AWS team followed Pintea and field researchers in Gombe National Park to better understand how they work in the field. The breakthrough, the team said, was that the tool could be used for the digitization of handwritten records as well.
The Jane Goodall Institute and AWS said the AI platform could be used for five tasks: multimedia search, video scene detection, chimpanzee facial recognition, AI-powered behavioral analysis, and automated data processing and translation.
By December 2025. AWS. the Jane Goodall Institute. and AI tool and data platform builder Ode PB held a kickoff meeting to transition the new system called the Gombe AI Research Platform. Throughout early 2026, the teams worked to map out platform requirements and researcher workflows where AI could be helpful.
Some researchers at the institute were already experimenting with the platform. and Rashid said it is expected to be fully live by the fourth quarter of 2026. The teams are still developing elements including a specialized dictionary derived from 65 years of field notes and the ability to understand a Gombe-specific dialect of Swahili.
Pintea said he expects the AI to help the institute catch up with its six-year backlog of data that has yet to be digitized. He also said he hopes the Gombe AI Research Platform will be used by field researchers, principal investigators, Ph.D. students, and undergraduates beyond the Jane Goodall Institute.
The work lands far beyond one database. Pintea pointed to ongoing research in Uganda’s Kibale National Park. where two groups of chimpanzees have been fighting in the second-ever recorded civil war among the species. He said the institute’s AI efforts are making real-time field data more readily accessible to researchers studying the phenomenon. which was first recorded in Gombe.
It comes back, in the end, to what the institute has pursued for generations: “It’s about understanding our closest living relative, as a way to understand human evolution and define ourselves,” Pintea said.
Jane Goodall Institute chimpanzee research Gombe AI Research Platform Amazon Web Services AWS Imagine Conference large language models digitization handwritten notes WISE Gombe National Park Kibale National Park conservation science