GenAI Builder

Overview

SnapLogic’s GenAI Builder enables you to build Language Learning Models (LLM) powered custom co-pilots based on your use cases. Use this no-code solution to accelerate and automate your business workflows.

This article describes how to use GenAI Builder to build your AI application, which can be used to answer employees' HR-related queries, review legal documents, summarize companies' quarterly results, and more.

Support Matrix

Component Snap Pack Vendor Description
LLMs Amazon Bedrock LLM Snap Pack Amazon Bedrock
  • Integrates with diverse Amazon's Foundation Models (FMs)
  • Supports prompt generation, chat completions and embeddings
Azure OpenAI LLM Snap Pack Azure OpenAI
  • Integrates OpenAI's language learning models (LLM) with Azure OpenAI Service
  • Supports prompt generation, chat completions and embeddings
OpenAI LLM Snap Pack OpenAI
  • Integrates with OpenAI's language learning models (LLM) within the SnapLogic Platform
  • Supports prompt generation, chat completions and embeddings
Vector Databases Pinecone Snap Pack Pinecone
  • Queries records in the Pinecone database
  • Processes the vector from the input document
MongoDB Atlas Vector Search MongoDB
  • Performs similarity searches
  • Supports Approximate Nearest Neighbor (ANN) queries
  • Supports Range queries
Utilities LLM Utilities Snap Pack -
  • Enables custom knowledge-context chunks with defined properties.
  • Facilitates embed and store result chunks in the vector database to optimize data organization in subsequent queries
HTML Parser - Parses HTML into text
Markdown Parser - Parses Markdown into text
HTML to Markdown Converter - Converts HTML into Markdown

Prerequisites

Summary steps

  1. Configure your Amazon Bedrock Account, Azure OpenAI API Key Account, OpenAI API Key Account, and Pinecone Account in SnapLogic.
  2. Launch the Indexer and Retriever pattern pipelines from the Public Pattern Library.
  3. Modify the Snap Settings of the pattern pipelines' based on your requirements.
  4. Validate and test the pipelines.
  5. Launch the chat interface and prompt.