Deasie
Deasie is a B2B analytics company from the Y-Combinator S23 batch, specializing in data governance for language models, ensuring the use of safe, high-quality, and relevant data.
Company Overview
Deasie is a B2B analytics company that focuses on data governance for language models. The company was part of the Y-Combinator S23 batch and employs a small team of three. They provide data governance layers to ensure that only safe, high-quality, and relevant data is fed into language models used by enterprises, thereby enhancing the reliability and safety of AI applications.
Platform Features
Deasie has developed a platform that connects to thousands of documents, automatically tagging each with metadata related to its content and sensitivity. Key features include auto-generating metadata, identifying and removing sensitive information, and controlling data access. Every document is chunked into smaller text fragments based on semantic meaning and labeled with metadata, including contents within charts and tables. This ensures that AI models only receive relevant and safe data, enhancing the reliability of language model applications.
Industry Recognition
Deasie has gained industry recognition for its innovative approaches in measuring data quality and relevance for unstructured data. Notably, TechCrunch has highlighted the company for its pioneering work in this area. The platform's ability to label data, find sensitive information, and control data access has been recognized as novel and effective, making it a valuable tool for enterprises.
Investors and Funding
Deasie is backed by prominent investors, including General Catalyst, Y Combinator, RTP Global, and several world experts in enterprise data. These investors support the company's mission to provide robust data governance tools for language models, reinforcing the reliability and safety of AI applications in enterprise settings.
Career Opportunities
Deasie employs a remote team and is actively expanding. The company is currently hiring for various positions, including the role of Founding Engineer (Full-Stack) in New York, NY. They are looking for talented individuals who are interested in contributing to their innovative platform and enhancing the safety and reliability of language models with high-quality data governance tools.