DBbun — Turn Any Content into Interactive Simulators

Upload any document, image, or file — research papers, patents, news articles, technical reports, and more — and receive a complete simulation bundle with runnable code, synthetic datasets, and documentation. All outputs are fully synthetic, generated using AI-driven methods to support analytics, education, machine learning, invention review, technical due diligence, and exploration.

• Explore Patent Demo
• View demos on IEEE DataPort
• Browse datasets on Hugging Face
• Watch demos on YouTube

What is DBbun for?

Research and education — teaching machine learning, statistics, analytics, and scientific reasoning using fully synthetic data and executable companions.
Prototyping and validation — exploring ideas, workflows, and pipelines before access to real-world data is available.
AI development and testing — experimenting with models when real data is unavailable or restricted.
Patents and intellectual property — turning patent documents into more explorable technical companions, including synthetic data, simulators, visual outputs, and related assets that can help with invention understanding, technical review, comparison, and communication.
Government and innovation programs — simulation, stress-testing, and evaluation of analytic methods in controlled environments, including contributions to open innovation programs such as the DARPA Lift Challenge.

Founder Background

DBbun was founded in September 2025 by Uri Kartoun, a data scientist, inventor, and PhD in Intelligent Systems with over 15 years of experience in real-world evidence, predictive modeling, and large-scale data solutions at Microsoft, IBM, and Harvard/Mass General Hospital. Uri is the author of 85+ patents and has developed pioneering methods for generating and analyzing complex datasets.

Publications & Recognition

DBbun is an active, SAM-registered U.S. small business (UEI: QY39Y38E6WG8; CAGE: 16VU3). DBbun completed the NSF I-Corps program through the MIT / New England Node as part of the February 2026 cohort. In April 2026, this broader vision was reflected in a Communications of the ACM Letter to the Editor by founder Uri Kartoun, titled “AI and the Evolving Role of the Scientific Paper,” published alongside a response from Vint Cerf.

History & Inspiration

During his fellowship at Harvard/Mass General Hospital, Uri created EMRBots, a non-profit project that generated synthetic EMR-like data long before generative AI became popular.
EMRBots became widely used in teaching and research.
It inspired development of a new type of neural network.
Its popularity and impact on the scientific community laid the groundwork for DBbun.
Some of the core ideas later appeared in a 2018 article in Communications of the ACM, which explored how representative synthetic datasets can enable research when real data is unavailable.

Disclaimer

All DBbun datasets are generated from public-domain information or fully synthetic, imaginary data, and no real patient or personally identifiable information is used. They are intended solely for research, teaching, prototyping, and analytics, and are not suitable for clinical decision support or direct patient care. Users are fully responsible for any use or application of the data. Uploaded files are processed in memory and are not stored, retained, or reviewed by DBbun. No copies of user-uploaded content are kept after processing is complete. References to DARPA programs are for descriptive purposes only and do not imply endorsement, funding, or affiliation by the U.S. Department of Defense or DARPA.

Intellectual Property

DBbun owns utility patent-pending and trade-secret technologies that support its unique approach to synthetic dataset generation.