Tech Tips: 2024

Thursday, February 22, 2024

Redis RAG with Nvidia NeMoGuardrails

Summary

This post will cover the usage of guardrails in the context of an RAG application using Redis Stack as the vector store.

Nvidia's guardrail package is used for the railed implementation.
Langchain LCEL is used for the non-railed implementation.
Content from the online Redis vector search documentation is used for the RAG content
GUI is implemented with Chainlit

Application Architecture

This bot is operating within a Chainlit app. It has two modes of operation:

'chain' - no guardrails
'rails' - NeMo guardrails in place for both user inputs and LLM outputs

Screenshots

Bot without rails

This first screenshot shows the bot operating with no guardrails. It does just fine until an off-topic question is posed - then it cheerfully deviates from its purpose.

Bot with rails

Same series of questions here with guardrails enabled. Note that it keeps the user on topic now.

Code Snippets

Non-railed chain (LCEL)

Railed with NeMO Guardrails

Source

https://github.com/redis-developer/llm-redisrail

Monday, January 15, 2024

Change Data Capture w/Redis Enterprise

Summary

Redis Enterprise has the capability for continuous data integration with 3rd party data sources. This capability is enabled via the Redis Data Integration (RDI) product. With RDI, change data capture (CDC) can be achieved with all the major SQL databases for ingress. Similarly, in the other direction, updates to Redis can be continuously written to 3rd party targets via the write-behind functionality of RDI.

This post covers a demo-grade environment of Redis Enterprise + RDI with ingress and write-behind integrations with the following SQL databases: Oracle, MS SQL, Postgres, and MySQL. All components are containerized and run from a Docker environment.

Architecture

Ingress

Write-behind

Code Snippets

Docker Compose - Redis Enterprise Node

Docker Compose - Oracle Enterprise

Docker Compose - Debezium

RDI Ingress w/Prometheus Integration

Source

https://github.com/Redislabs-Solution-Architects/redisdi-docker

Monday, January 8, 2024

Document AI with Apache Airflow

Summary

In this post, I cover an approach to a document AI problem using a task flow implemented in Apache Airflow. The particular problem is around the de-duplication of invoices. This comes up in payment provider space. I use Azure AI Document Intelligence for OCR, Azure OpenAI for vector embeddings, and Redis Enterprise for vector search.

Architecture

Code Snippets

File Sensor DAG

OCR DAG

OCR Client (Azure AI Doc Intelligence)

Embedding DAG

Embedding Client (Azure OpenAI)

Vector Search DAG

Vector Search Client (Redis Enterprise)

Source

https://github.com/Redislabs-Solution-Architects/docai_pipeline

Monday, January 1, 2024

Redis Search and SQL Command Comparison

Summary

This post covers comparisons of various data search scenarios of equivalent SQL and Redis Search commands. The Chinook data set is deployed in Oracle Enterprise. The Oracle Enterprise data is then continuously populated into Redis JSON objects using Redis Data Integration (RDI).

Architecture

This entire architecture is deployed in Docker containers.

Sample Scenarios

Scenario - Which countries have the most Invoices?

SQL

Redis Search

Scenario - Which artists have written the most Rock music?

This particular query touches 4 different tables in relational db tables. There are 2 approaches to this in Oracle:

Ad hoc query that implements the necessary joins
Materialized View that implements the same join query but is updated continuously.

SQL - Query

SQL - Materialized View

Below is the equivalent Materialized View. The ALTER commands set this view up to be tracked by Debezium and ultimately replicated in Redis Enterprise.

Redis Search - Materialized View

The Oracle Materialized View is treated the same as any other table by Debezium and is thus populated into Redis via RDI.

Redis Search - Triggers & functions

An alternate approach to the materialized view is to perform the necessary joins in a Redis Function. This a Gear 2.0 feature where a function can be written in Javascript and executed server-side in Redis. Snippet below of that function.

Source

https://github.com/Redislabs-Solution-Architects/search-sql