Composable, data and AI building blocks, including models for AI predictions, and a petabyte-scale cloud data platform preloaded with 100TB+ of Web3 data. Create, fork, and share Datasets and Models with the new GitHub-integrated Community Edition.
Hi, All đź‘‹
Since the Waitlisted Preview launch of Spice.ai last year, we've worked with projects like Yakoa, EigenLayer, and Entendre Finance who have leveraged building blocks from Spice AI’s platform to create high-performance, highly available, data and AI-driven applications.
Today, we’re announcing Spice.ai is publicly available for everyone, including a brand-new community-centric hub and a new Community Edition plan complimentary for developers!
Check it out at https://spice.ai
With Spice.ai, you can:
- Query real-time and historical data using SQL in seconds and fetch results in JSON or Apache Arrow for use with applications, ML, or libraries like NumPy and Pandas.
- Collaborate on datasets, ML models, and applications in one place through the new community-centric developer hub.
- Create, fork, and curate your own hosted datasets and ML models to share and build with the community.
- Define and manage custom datasets in GitHub along with your application code.
- Bring your own data or leverage the 100TB+ of preloaded Web3 and Prices data.
Spice.ai is preloaded with community data from ecosystems including Ethereum, Bitcoin, Polygon, and EigenLayer along with enriched datasets for NFTs, DeFi, DEXs, ENS, Prices, and more. Most datasets are open-source and available at github.com/spiceai/datasets.
We're also launching:
Spice Firecache: Turbocharge Your SQL Queries
Spice Firecache is a real-time, in-memory SQL service based on cloud-scale DuckDB instances that enables blazing fast SQL query up to 10x the performance of general SQL query. EigenLayer uses Spice Firecache to enable scenarios not possible before, including serving dynamic data to their high traffic dashboards, real-time monitoring, and analytics. ML models hosted on the Spice.ai platform can be paired with Firecache to power fast, low-latency inferencing, as demonstrated by the AI predictions demo on the Spice.ai website.
If you like DuckDB, you’ll love it at cloud-scale, automatically provisioned and updated with real-time data.
Spice Functions: Beyond SQL
Spice Functions is a hosted Bacalhau service for serverless function compute that enables data transformation and processing co-located with your data and machine learning pipelines. Developers can write code in Go or Python. And every Spice Function deployment gets access to a dedicated hosted DuckDB instance for working data.
Spice Functions unlocks scenarios that are difficult, expensive, or even impossible in pure SQL, such as adding to a basic accumulator or applying a custom algorithm on each new block of data.
Custom Datasets and Views: Your Data, Your Rules
Developers can tailor Spice data to their application with Custom Datasets and Views, defined in GitHub, alongside their application code. Datasets can be populated with Spice Functions and by connecting to external data sources starting with PostgreSQL. We’re excited to announce that Yakoa — IP protection for the blockchain — is one of the first projects to make their NFT data available in the Spice.ai platform with the release of their Copymint datasets.
“Other solutions were prohibitively expensive — what we could do in Spice with a single query would have taken millions of API calls in other platforms.” - Andrew Dworschak, CEO & Co-founder of Yakoa
We’re also making over 15 EigenLayer datasets public that ecosystem participants can use to build data-driven experiences, like Nethermind has done with their Restaking Dashboard. Combined with Spice.ai’s rich Ethereum Beacon chain data and HTTP API, observability into the EigenLayer universe has never been easier.
Spice ML Models: Automated Machine Learning
Like Datasets and Views, ML Model definitions are synced from GitHub to the Spice.ai platform and connected to Spice.ai hosted data, hosted machine learning pipelines, and Spice Firecache. The entire machine learning data lifecycle from origin to processing to training and inferencing is automatically and seamlessly managed by the Spice.ai platform so that developers can create decision-making applications, such as predicting and responding to resource requirements or mitigating potential security concerns with ease.
Spice.ai