Running Shared AI Models with Avgidea Data Platform and YugabyteDB
summary
Avgidea Data Platform (ADP) demonstrates a one‑stop pipeline that pulls rows from YugabyteDB tables or S3, executes Python functions (pandas dataframes, preinstalled TensorFlow and PyTorch), scales across nodes, and writes CSV or table results back to YugabyteDB. Live examples include k‑means clustering that outputs image files and logs, and an RFM customer segmentation pipeline that reads transactions, clusters customers, and writes an mvcs table. An advertising flow pushes targeting lists to Twitter and shows measurable follower/engagement uplift when switching lists.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:06Data platform requirements
- 3:57ADP and YugabyteDB fit
- 5:22Data exchange features
- 10:09AI model execution
- 13:32Advertising features
- 17:34Case studies
- 20:07YugabyteDB integration demo
- 23:59Summary
related talks
37mHow YugabyteDB Empowers AI-powered Fraud Intelligence
26mEfforts Toward Adoption and Uncovering Hidden Features
25mDiscover the Easiest Path to Distributed DBaaS: Yugabyte Cloud
28mDisrupting the Status Quo with Distributed SQL to Deliver Data-Driven Solutions
34mAnalytics and AI with PostgreSQL - Mission Possible!
38mExploring Azure Databricks Notebooks & Power BI for YugabyteDB
49mBuilding Advance Analytics pipelines with Azure Databricks
24mPostgres-powered AI: Running an End-to-End AI Platform with Postgres on Azure