Introducing Spice AI Flow

Margaret Zhang
The Spice AI Blog
Published in
4 min readJan 30, 2023

--

Spice AI Flow is now available in Preview!

Spice AI Flow is your database query superpower. It autocompletes SQL queries in realtime and suggest datasets and tables directly in app.

Spice AI Flow generated query recommendation

At Spice AI, we’re on a mission to support developers with the fastest, easiest and smartest way to build data and AI-driven apps in Web3. Currently, Spice AI provides one of the fastest query times across indexed datasets in Ethereum, Bitcoin, and Polygon. However, we’ve noticed that developers still spend considerable time composing and editing queries in SQL — time they can spend tackling complex coding challenges with creative solutions.

We believe that commercially available AI technology has already fueled leaps in developer productivity. By embedding the autocomplete extension into Spice’s query editor, we remove SQL nuances that distract developers from their workflow, effectively reducing their overall time-to-results.

Combining Spice AI Flow with Spice’s lightning fast querying speeds, now you can truly query blockchain data within seconds.

Spice AI Flow generated query recommendation in action

Anyone can now query with plain text, hit enter, and watch as Flow autocompletes the request with the corresponding SQL lines in your spice.xyz query editor. Users can then hit tab to accept the AI-generated completion and accept, reject, or edit before letting it run.

This brings SQL-novices access to blockchain data with precision, and enables developers already familiar with SQL to query common requests even faster.

See the step-by-step guide and query examples in the Spice Docs here →

SQL Query Standard Completion

We leave to you the option to turn off autocomplete and manually refine complex queries. Even when Flow is turned off, you will receive non-intrusive inputs suggestions (column and table names, keywords) to help rapidly complete lines of SQL query as you type. This directly reduces query writing errors, gently reminds you of the correct inputs, and provides a range of related datasets to explore.

SQL Query Editor suggests the available fields for eth.recent_blocks along with their type.

See the step-by-step guide and query examples in the Spice Docs here →

Spice AI Flow enables Web3 Developers to

Query with Confidence: Partner with SQL prompts you can count on and edit, accept, or remove suggestions before you run your intended query.

Stay in the Flow: query error free, get input suggestions, and discover related queries all without leaving the editor to stay focused with ease.

Take time back to focus on what matters most: use the time saved to learn SQL writing from Flow’s generated queries OR spend it on your most pressing coding challenges (OR take a much needed break).

Spice team members already rely on Spice AI Flow in our everyday work so we’re thrilled to see more web3 developers embrace (and get used to!) querying and app-building experiences that flow.

About Spice AI

Spice AI is the fastest, easiest and smartest way to build data and AI-driven apps in Web3. Designed for applications and machine learning, we remove the complexity of building and operating massive data infrastructure — enabling teams to create the next generation of data and AI-driven apps.

We are the first service to deliver developer-friendly bulk data over high-performance Apache Arrow APIs. We’ll soon be the first to deliver AI/ML capabilities for blockchains.

If you’re developing on Polygon, Ethereum, or Bitcoin, you can build effortlessly with Spice’s enterprise scale data and AI infrastructure in just three lines of code.

Get started now →

👥 Join our new and growing Community:

Twitter | LinkedIn | GitHub | Discord | Telegram

The Team at Spice.

Have questions or feedback? Contact us

Spice.ai | View previous releases

--

--

Head of Partnerships @Spice AI | Ex VC Fellow @INVC (M12, Notion) | PM @Gartner | Investment Summer Associate @Temasek | Amateur Mountaineer | NYU Stern Alum