The Training Data Project (TDP) was established in 2023. Founded by practitioners and entrepreneurs, our roots are in AI/ML, the data labeling industry, and in serving and in supporting the Defense, Intelligence, and Public Safety community for over 25 years. We have lived the pain of proprietary platforms, minimal standards, lack of interoperability, closed architectures, and needless expense in fixing errors. As experts, citizens, and taxpayers- we are certain there is a better way in training data.
Our mission is to change training data from the ground up. TRUST defines our mission at TDP. We strongly believe that transparency, ready-access, quality, standards and traceability in data labeling are essential for AI to be effective in today’s world. Efficient and effective management of data labeling and training data across both public and private sectors is crucial for driving AI/ML innovation. Public and private sector organizations must have access to information, tools, and resources for informed decisions about the training data that drives their AI/ML initiatives.
We firmly believe that TRUST-worthy AI training data is a national priority. Everything in a data label should be measurable, and everything in a data label should be a choice. Through the work of our members and partners, and through collaboration with academia and other nonprofit organizations, we aim to provide TRUST-worthy advice in adopting and implementing AI for the public good.
Get in touch with us to get help with access to open source tools, best practices in labeling, governance, how to choose your training data for model training, and other questions. Follow us for new programs and resources as TDP continues to grow and expand.
