DATS-4 Database AI Suite

About the App

User Guide

The DATS-4 Database AI Suite is a tool for AI-driven data analysis against relational databases. Its primary function is to connect to remote PostgreSQL or MySQL databases to execute queries. Queries can be simple or part of a complex analysis, run through either a single AI agent or a multi-agent framework.

The tool also supports file uploads (CSV, tab-delimited). Uploaded files can be loaded into a pre-existing database connection or into a temporary database created on the fly. Additional features include the ability to generate Python charts, perform statistical analysis, and create PDF reports.

Quick Start 1: Using the Sample Button

The fastest way to get started with REX-A Analyzer is through the Sample button feature:

1. Access Sample Files:

   • Click the "Sample" button in the top-right corner
   • Browse through curated sample datasets with detailed information about size, row count, and columns

2. Select and Analyze:

   • Choose any sample file that interests you
   • Each sample comes with:
     • File description
     • Size and structure information
     • Pre-crafted analysis prompts

3. Database Options:

   • After selecting a file, choose between:
     • Temporary Database: Instant analysis environment (Recommended for quick start)
     • Your Own Database: Connect to your existing database

4. Quick Analysis:

   • Copy the provided analysis prompt (Click "Copy Prompt" button)
   • Navigate to the AI Advanced Analyst tab
   • Paste and customize the prompt for detailed analysis

💡 Pro Tip: The sample files are carefully curated to demonstrate different analysis capabilities, from simple statistical analysis to complex data transformations.

Quick Start 2: Sample Files on Google Drive

Sample files are available on Google Drive for quick testing, including:

• RBI Cards / ATM / POS live data (small but rich datasets)
• Mock customer profile datasets (100K to 10M rows)
• Various file sizes from 10MB to 1.2GB

Access Sample Files

Interface Guide

Tabs Overview

• AI Data Structure: AI interpretation of file structure with PDF report
• AI Quick Insights: Quick analysis of uploaded file using 100 rows of data with PDF report
• AI Quants Analyst: Portfolio and Quants reporting agent with addtional databae and python capabilities
• AI Database Analyst: Main Database AI agent. Sequential agent with multi-step analysis - reasoning, execution, and error handling
• Logs: Processing logs

Menu Options

• Sample: Quick start with sample files

• Choose File: Upload CSV / TXT files

• Fast Insights → Table: Load uploaded files into interactive tables

• Fast Insights → Structure: AI analysis of uploaded file structure using sample rows

• Fast Insights → Analysis: AI analysis of uploaded file using 100 rows of data

• Connect to DB: Connect to your own warehouse using credentials

• Upload File to DB: Upload files to your connected warehouse (temporary database available without login)

• Export Data: Export files from connected warehouse

• Create DB: Create Temporary Postgres Database (requires login)

• Model: Select the AI model and agent for Advanced Analyst

Features & Video Guide

1. BYOW (Bring Your Own Warehouse)

• Connect to MySQL and PostgreSQL warehouses
• Analyze existing warehouse data
• Simple credential-based connection
• Unlimited table sizes and query processing
• Operations executed on warehouse side
• Supports petabyte-scale data across thousands of tables

2. Natural Language-to-SQL Querying

Video Guide

The video guide is for REX-3 (Version 3 build ) but would be helpful for DATS-4 too. Comprehensive Guide: Advanced Analysis with Sequential Reasoning

Video Timeline:

• 00:00:00 - Capabilities Overview
• 00:02:08 - DB Connection & Analysis
• 00:06:43 - File Upload & Analysis
• 00:09:21 - Sequential Agent Framework
• 00:15:02 - Performance Considerations
• 00:23:09 - Cost Considerations
• 00:29:49 - Live Error Debugging
• 00:37:37 - Architecture & API Flows
• 00:45:08 - Deployment Guide
• 00:58:19 - App Functionalities
• 01:01:05 - Resources
• 01:01:51 - Conclusion

Example Queries

• "Pull up record for Customer ID 12345"
• "Show response rate distribution by Education and Housing"
• "Display response rate by 10-year age buckets"
• "Analyze customer demographics with specific filters"

3. Interactive Tables

• Sort, filter, and search functionality
• Row details popup via calculator icon
• Column statistics with distribution metrics
• Comprehensive statistical analysis tools

4. File Analysis Features

• Support for TXT/CSV files (pipe or comma-delimited)
• Structure analysis with AI
• Detailed data analysis options
• File size capabilities:
  • Tested up to 600MB (5M rows)
  • App limit: 1.5GB (customizable)
  • Performance varies by file size
  • Interactive features work best with files under 10MB

5. Warehouse Integration

• Push data to your warehouse
• AI-powered data analysis
• Automatic schema detection
• Interactive query capabilities

6. Advanced Analytics

• Python-based statistical analysis
• Custom data transformations
• Chart generation
• Support for both warehouse and uploaded data

Source Code and Build Guides

Build Guide

The full app has 7 major components:

1. Main App: The main application with the UI
2. FastAPI Server:Database Connector: Handles Text-to-SQL processing, including file uploads.
3. FastAPI Server:Neon DB Creation: Handles Text-to-SQL processing, including file uploads.
4. LLM Agent: Sequential Agent Framework with LLM Agent built with Flowise AI.
5. Proxy Server: For API Calls to OpenAI / Gemini / Openrouter
6. MCP Server - Markdown to PDF: For converting markdown to PDF
7. Quant Analyst: This is the TIGZIG Quants Agent app integrated into a single tab to provide an integated experience to users.

GitHub Repositories

1. Main Repo
   https://github.com/amararun/shared-rexdb-auth-embed-v3-agentflowv2

2. FastAPI Server : SQL DB Connect
   https://github.com/amararun/shared-fastapi-rex-db-coolify

3. FastAPI Server: Neon Database
   https://github.com/amararun/shared-rexdb-fastapi-neon

4. Sequential Agents Schema - Flowise
   In docs folder in Main Repo, including JSON for Database Connect Tool.

5. FastAPI : LLM API Calls Proxy Server
   https://github.com/amararun/shared-rtWebrtc-fastAPI-ephemeral

6. MCP Server: Markdown to PDF
   https://github.com/amararun/shared-mcp-markdown-to-pdf

7. Quants Analyst: Head over to the Quants Agent App from main menu and refer corresponding docs for source codes for the backend. Flowise JSON schema available in the docs folder

Detailed Build Guide

The video guide is for REX-3 (Version 3 build ) covering 80% of the DATS-4 build process. However the GitHub repos are latest for Version 4 with updated Readme with latest build instructions.

YouTube Video
YouTube Video Guide:
https://www.youtube.com/watch?v=hqn3zrdXVSQ

00:00:00 - Quick Overview of Capabilities
00:02:08 - Connect to DB & Analyze: Modeling Data Mart & Customer Profile Summary
00:06:43 - File Upload & Analyze: India Bank Ranking - Credit Cards
00:09:21 - Sequential Agent Framework: Setups & Orchestration
00:15:02 - Performance Considerations (Quality, Speed, Reliability, Latencies, Agent Backend, SQL Call Failures, Database Server, API Call Failures, Error Catching, Validations)

00:23:09 - Cost Considerations
00:29:49 - Live Error Debugging
00:37:37 - High Level Architecture & API Flows
00:45:08 - Deployment Guide & GitHub Repo Walkthroughs
00:58:19 - App Functionalities - How to Have Cursor Explain it.
01:01:05 - Top Resources
01:01:51 - End