According to Yahoo News, Retool, a platform for building line-of-business apps, is launching new tools to help users build AI-based apps, including a hosted vector store for adding context to large language models (LLMs). Retool CEO and co-founder David Hsu explained that many customers are exploring AI integration in their apps, but most enterprises find value in reasoning over their internal data. Fine-tuning existing models with a reasonable amount of data is possible, but using all of a company's production data is not feasible, and the data would soon be outdated.
To make custom data easily accessible for LLMs, the state-of-the-art method is to vectorize the data. Companies like Google, Microsoft, DataStax, and MongoDB have all launched vector search services in recent months. Hsu believes the main challenge for Retool's customers is getting data into a vectorized database and keeping it up to date. To address this issue, Retool is launching Retool Vectors, a hosted vector storage service using the open-source pgvector extension for Postgres at its core.
Retool has also launched AI-based actions for common use cases like text summarization and classification, image generation, and more. The company partnered with OpenAI for these features, which includes integrations into Retool Workflows. Brad Lightcap, Chief Operating Officer at OpenAI, expressed excitement about the partnership, stating that tools like Retool help businesses put AI into production much faster without compromising safety.