On SQrL. Q&A with Arnaud Comet
Q1. You have launched a new feature called SQrL. What is it?
SQrL is a bot we created that is trained on SingleStore specific content from documentation, to GitHub, and other sources. Engineers and others can use it to write and optimize SQL, generate code blocks and build real-time applications.
Q2. Is SQrL different from ChatGPT-4? If yes, what are the differences?
Yes. SQrL is a bot that we created specifically for SingleStore related content. While ChatGPT is a powerful language model that can be used for a variety of tasks, we think it’s even more powerful to create bots like SQrL that use custom data to improve accuracy and speed.
Q3. How can SQrL help anyone using SingleStoreDB to discover answers on their own data in real-time?
SQrL serves as a co pilot for anyone using SingleStoreDB, they can use it to write and optimize SQL, generate code blocks and even build real-time applications.
Q4. SingleStoreDB is a real time analytics database. Can you give some examples on how is it possible to query SingleStoreDB using SQrL?
Just ask a question to SQrL on how to write a query with specific context and SQrL. We do not support today metadata context for security and governance around our customer data and metadata.
Q5. SQrL is a ChatGPT bot and not a query language. Am I correct?
Q6. How do ensure that the answers are accurate and in real-time?
You just need to copy and paste the answer into the notebook or SQL editor and run it. If it doesn’t work, just tell SqrL what the error message is and it will be able to correct itself.
Q7. What are the steps to be taken to start using SQrL?
SingleStore users simply go to https://www.singlestore.com/sqrl/ and type in questions.
Q8 Anything else you wish to add?
There are plans to add on to SQrL to make it even more robust in the future including:
- Notebook integration where SQrL will provide code and help debug directly into the cells
- Metadata integration to provide solutions based on customer metadata (will require customer opt-in)
- Provides E2E scenarios around optimizing workloads, workload migration or sizing workload (and more)
- [Longer term] Proactive recommendation from SQrL
Arnaud Comet is director of product management at SingleStore where he is leading product-led growth, developer experience, pricing and cloud profitability. Prior to joining SingleStore, Arnaud worked at Microsoft on Azure Cosmos DB and later was a founding member of Azure Synapse Analytics where he also led many integration initiatives such as Azure Data Explorer, Azure Synapse Link and Azure Purview.
Play Demo Video
Sponsored by SingleStore