According to a new press release, Tonic.ai has launched Tonic Textual, a secure data lakehouse for large language models (LLMs), designed to address integration and privacy challenges in leveraging unstructured data for generative AI. This platform aims to streamline the preparation of unstructured data for retrieval-augmented generation (RAG) systems and LLM fine-tuning, tackling significant obstacles in enterprise AI adoption. Tonic Textual converts complex and messy unstructured data into AI-ready formats, enabling seamless embedding, fine-tuning, or ingestion into vector databases while ensuring data security and privacy.

Unstructured data, which constitutes about 90% of enterprise data, holds significant untapped value but presents extraction and standardization challenges. Enterprises generate vast amounts of unstructured data, which is often siloed and manually processed, hindering AI deployments. Tonic Textual addresses this by automating data extraction, transformation, and protection, allowing developers to focus on data science rather than data preparation. The platform’s advanced capabilities ensure sensitive information is identified and safeguarded, mitigating compliance risks and enabling responsible AI utilization.

Tonic Textual provides a comprehensive solution for managing unstructured data, offering features like sensitive data redaction, enrichment with metadata, and automated data pipelines. The platform supports various data formats and plans to expand with native SDK integrations, enhanced data management tools, and broader data connector libraries. By simplifying the complexity of handling unstructured data, Tonic.ai empowers enterprises to leverage their proprietary data for AI initiatives securely and efficiently, supporting the company’s mission to balance innovation and privacy.