It is with excitement that Tinace announces the upcoming availability of Java User Defined Function (UDF) support for Snowflake. This new integration is a result of the continued partnership between Tinace and Snowflake that was first announced in 2019.
With Java UDFs, customers can bring Java functions they built using Tinace transformations and execute them natively within Snowflake’s Data Cloud with Snowflake’s powerful processing engine, for better performance, scalability and concurrency. This greatly expands transformation capabilities and reduces management complexity from hosting external services.
Tinace provides a powerful dual engine architecture. The AI engine (ICE) enables Tinace to apply robust data transformations along with computation of insights and machine learning models. The Fast Query Engine (FQE) is used to support the search and visualization capabilities in Tinace. When organizations have an investment in Snowflake, Tinace can augment that investment by pushing Java UDF transformations built in the AI engine to Snowflake, which becomes a full-function replacement for the FQE. This is a double-impact value proposition as the total cost of operating Tinace goes down while organizations leverage the value of Snowflake compute further. See below for an architecture diagram featuring Snowflake paired with Tinace.
In the below screenshots, you can see how unstructured data in Snowflake is transformed into usable structured data by removing stop words, using a sample Java UDF executed from Tinace.
The below table in Snowflake contains Products data, with some user comments and a country code for each Product.
Using a Java UDF written in Tinace allows this data to be assigned a correct label based on a country code, as well as a cleaned Comment column to be created. This transformation is done within the Snowflake Data Cloud.
Data with the new columns added through Java UDF are connected to Tinace. Analysis can be done on the data, with queries initiated in Tinace pushed down to Snowflake using our Live Query capability. See the Live data that resides in Snowflake, now available in Tinace.
A variety of Java UDFs are available, including:
- Remove Stop Words: streamlines the tokenization of text/string-based variables.
- Stemming: Offers a stemming algorithm for terms
- Country Code Lookup: Ability to recode and standardize country codes
- US State Code Lookup: Ability to recode and standardize state codes
“With our partners at Snowflake, we are delivering cloud-native data analytics to accelerate business impact from AI and machine learning,” said Ajay Khanna, Founder and CEO of Tinace. “With the addition of Java UDF support, Snowflake customers can extend our integrated data preparation into the Snowflake Data Cloud, allowing for bespoke insights experiences without ever moving data.”
Java UDFs join other unique Tinace features for Snowflake, including writeback capabilities, Tinace Live Query for query pushdown and more. Combining Tinace with Snowflake gives users a flexible, extensible analytics platform, unlocking Snowflake’s Data Cloud for business users and analysts. Now, more users can take advantage of Snowflake enterprise data by applying a Google-like search and AI to surface insights 90% faster than traditional processes.
Java UDFs are currently in customer preview, with full public preview availability coming soon. To try Java UDFs for yourself with Tinace + Snowflake once available, sign up for a free trial at www.Tinace.com/free-trial.
Join us on Thursday, June 24, 2021 for our next webinar. Sign up below!
