ML meets NoSQL: Integrating Python user-defined functions (UDFs) with N1QL for analytics
summary
Couchbase Server 7.0 (Developer Preview) adds python UDFs to Analytics and external data access from AWS S3/Azure Blob so queries can join local JSON buckets (airport ~1,000 records) with external hotels and city details without ingestion. Demo scans ~30,000 documents and returns a three‑way join in ~3.69s on a 1GB laptop. Machine learning model is trained on Rotten Tomatoes, packaged with shiv and uploaded to the Analytics library, exposed as a UDF (model.getsentiment) returning rotten/fresh/certified fresh inline in N1QL.
AI-written summary · about
chapters (12)
- 0:00Intro
- 0:24Agenda
- 1:57Couchbase customers
- 2:33Couchbase platform
- 5:01New features
- 6:34Developer preview
- 7:37Benefits
- 8:43Challenge
- 10:23Creating the link to Azure
- 12:55Adding data to Azure blobs
- 15:05Creating an external analytics collection
- 19:17Analyzing external data
related talks
38mHow to get real-time insights from your JSON data?
40mData Intelligence Platform on AWS
14mMulticloud Analytics Made Simple
37mAccelerating Python Analytics by In-Database Processing
29mBlazing-Fast Python in Your Database: Unlocking Data Science at Scale with Exasol
51mSpeed, scale, query: can NoSQL give us all three?
59mData Modeling for
21mUse cases that drive the need for NoSQL databases