August Homeruns

Author

Varun Mishra
August 19, 2024

Did you know that “Good Morning” in Japanese is おはようございます (ohayoo gozaimasu) when used in a formal setting. It is おはよう (ohayoo) if you are using it in an informal setting.

The data and analytics world was filled with announcements of AI integrations and Intelligent automation. Here is August’s espresso shot of top features brewing in the world of Business Intelligence and Data Engineering:

Business Intelligence

Google launched the Summarization extension in Looker offering Vertex AI’s Gemini model to summarize your dashboard data. It now provides capabilities of Summarization, Prescription and Action!

Why it matters: While such summarization features are available across the board with BI platforms, Looker has integrated Prescription capability which allows for next steps based on your data!

Tableau launched Einstein Copilot to prepare data sources, create visualizations, and tell stories using Generative AI. Einstein Copilot is built on Salesforce’s Einstein Trust Layer so inherited features for security, governance, and trust capabilities come with it.

Why it matters: Generative AI is breaking down barriers to moving across BI products. Features like Einstein make it easier for non-Tableau users to enter into the world of Tableau with ease.

Data Engineering

AWS has launched Anomaly detection in AWS Glue (the serverless integration service - think Azure Data Factory). The Anomaly Detection algorithm learns and predicts variances in Data Quality to avoid having to do manual legwork when Business Conditions change.

Why it matters: If you have had to ever go back to your ETL jobs to update load conditions to exclude certain values, which are no longer relevant, you would know how cumbersome it could be. Anomaly detection could save time and dollars spent by automating conditions used in Data Integration.

(AWS)

Databricks SQL Serverless is now available on Google Cloud Platform which would now provide SQL Warehouse compute capabilities on a Lakehouse.

Why it matters: In the era of cost control, companies want to avoid vendor lock-ins as much as possible. While cloud providers are moving towards self contained end to end solutions for data analytics, they may still not be suitable for enterprise analytics. Databricks + GCP is one such option offering that flexibility.

Snowflake has launched Cortex Analyst in Preview built using Meta’s Llama and Mistral models. Snowflake claims a 90% Text to Accuracy score for Cortex Analyst with GPT-4o coming in at 51%. Currently it can be integrated using Cortex Analyst Rest API.

Why it matters: AI in BI is a field that is heating up with Copilot (Microsoft), Datarooms (Databricks) and not Cortex claiming stake to provide solutions that help business users ask questions of their data in natural language using an interface.

Google has introduced Bigquery Continuous queries to analyze real time streaming data using SQL (emphasis on real-time). The use case that stood out is an abandoned cart on a website and how continuous queries can send this data in real time, analyze contents and can be used to send reminders to recover a sale.

Cool Jobs in Data!

Thank you for reading! If you liked reading this newsletter please share it with your network.

Feedback? Email us at: [email protected]

Reply

or to participate.