Get an insider’s look at AI, SQL, and blockchain tech in The Graph Builders Office Hours #32!
TL;DR: Learn about a session of The Graph Builders Office Hours where Sam Green discusses the integration of AI and SQL in blockchain analytics, showcasing Agentc's capability to enhance data interaction through natural language processing.
In episode #32 of The Graph Builders Office Hours, viewers had the opportunity to delve into the intersection of AI, SQL, and blockchain technology, courtesy of Sam Green from Semiotic Labs.
The session, packed with announcements, product demos, and a Q&A, shed light on The Graph’s innovative strides toward refining data analytics in the blockchain sphere.
Key announcements and features
The episode kicked off with some announcements, including a call for beta testers for the time series and aggregation feature now live on The Graph. This feature promises to simplify computations within subgraphs, significantly enhancing the developer experience.
The host added a reminder about The Graph Builders newsletter, a valuable resource for builders within the ecosystem. The newsletter summarizes the latest updates and ensures the community stays informed.
Subgraphs, SQL, and beyond
Sam Green’s presentation was the highlight of the hour, offering insights into the transformative work his team has been focusing on. Here’s a breakdown of the critical topics Sam covered:
- The Evolution of The Graph: The Graph has significantly evolved since its inception, starting with solving the indexing problem. Sam drew an interesting parallel between data and oil, emphasizing the importance of processing raw data to unlock its value. Subgraphs were pivotal for The Graph, providing a means to process and make data accessible via APIs. (Video ⏱️ 1:12)
- Introducing SQL to The Graph: Semiotic Labs has been pioneering the integration of SQL within The Graph ecosystem. SQL, renowned for its utility in analytics, offers a more versatile approach to data analysis than GraphQL. This initiative aims to leverage the existing PostgreSQL databases of The Graph to enhance analytics capabilities without forsaking the strengths of GraphQL. (Video ⏱️ 3:09)
- Subgraph SQL and Substreams SQL: Sam delineated between two significant efforts: Subgraph SQL, which focuses on querying data collected during the syncing of subgraphs, and Substreams SQL, a novel approach that allows for more flexibility in data collection and transformation. (Video ⏱️ 5:23)
- Agentc and Natural Language to SQL Conversion: An intriguing part of the presentation was the demonstration of Agentc, a tool that converts natural language queries into SQL. This tool simplifies data query processes and opens up possibilities for more intuitive data interaction. Sam explained Agentc is an experimental platform to investigate using natural language to simplify SQL data queries on The Graph. (Video ⏱️ 8:45)
- Demo and Use Cases: Sam provided a captivating demo, showing how he used Agentc to analyze Uniswap transaction data. The demo highlighted the potential of combining SQL with natural language processing to streamline data analysis in blockchain analytics. (Video ⏱️ 12:04)
The session ended with a Q&A. Sam addressed questions about the technicalities of implementing SQL on The Graph and the challenges of ensuring accuracy and reliability in natural language to SQL conversions.
Looking forward
The innovations presented in this session demonstrate The Graph’s commitment to empowering developers and enhancing the blockchain data analysis experience. As these tools and features continue to evolve, the possibilities for builders in the ecosystem expand.
Looking ahead, Sam highlighted an exciting direction for The Graph: integrating AI directly into its ecosystem. This vision would eliminate the need for external AI tools, like GPT, for data analysis. By embedding AI capabilities within The Graph, we could have a more seamless and intuitive experience, making blockchain data more accessible and engaging.
💡 This article answers questions like:
- Can AI be used to automatically create SQL queries?
- Are new query languages coming to The Graph?
- What's the difference between subgraph SQL and Substreams SQL?
- What is Agentc and how does it convert natural language to SQL?
No Comments