Data Startup Space and Time Creates Chatbot Powered by ChatGPT for Database Querying
ChatGPT will enable users to query the decentralized data warehouse using natural language prompts.
Jul 11, 2023 at 2:00 p.m. UTC:format(webp)/cloudfront-us-east-1.images.arcpublishing.com/coindesk/UIZMZLFVU5GM7PI3JSEDMTNVRI.jpg)
/arc-photo-coindesk/arc2-prod/public/LXF2COBSKBCNHNRE3WTK2BZ7GE.png)
Decentralized data startup Space and Time, backed by Microsoft's venture arm and Hashkey Capital, has integrated a chatbot using OpenAI's GPT-4, the latest version of ChatGPT, so that developers can maintain and query databases with ease, the firm told CoinDesk on Tuesday.
Space and Time has created decentralized data warehouse with both on-chain and off-chain information such that the datapoints can be analyzed and used by smart contracts.
The data warehouse is powered by zero-knowledge proofs (ZKs), a type of cryptographic proof that can prove the validity of a statement while only selectively revealing any information about the statement itself. The startup has applied ZK proofs on data queries, which are commonly run by using Structured Query Language (SQL), a programming language designed for managing and searching in relational databases, which store datapoints that are related to each other.
Space and Time's technology, proof-of-SQL, "allows the data warehouse to generate a SNARK cryptographic proof of SQL query execution, proving that query computation was done accurately and that both the query and the data are verifiably tamperpoof," according to its website. Within Web 3 environments, this is particularly important because there are no trusted entities to ensure the accuracy of the data or the query.
The integration of the chatbot, called Houston, using OpenAI's ChatGPT will make it easier to maintain and query databases using natural language prompts, as opposed to having to write code.
"Space and Time users can generate SQL or Python scripts from prompts, ask natural-language questions about data and get back an accurate visualization of the answer and load in new datasets, all by simply conversing with our chatbot," said Scott Dykstra, CTO and co-founder of the startup.
Houston also presents the SQL code generated through a natural language prompt so that it can be examined, and creates visual representations of the query and the results, said Nate Holiday, the startup's co-founder and CEO.
:format(webp)/s3.amazonaws.com/arc-authors/coindesk/6c6de681-6eb3-45a9-bfe0-c357743b23f9.png)
Eliza Gkritsi is a CoinDesk contributor focused on the intersection of crypto and AI.