AI Chatbot for Mainframe Data

A secure, natural-language chatbot for querying Db2 for z/OS mainframe data with multiple frontend interfaces (React + Streamlit) and a Java-based security gateway.

Architecture Overview

Architecture V1

┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│   Frontend      │────▶│   Node.js API   │────▶│   Java DB2 API  │
│  (React/Streamlit)     (Port 3001)      │     (Port 8080)   │
└─────────────────┘     └─────────────────┘     └────────┬────────┘
                                                          │
                                                 ┌────────▼────────┐
                                                 │   Db2 for z/OS  │
                                                 │   (Mainframe)   │
                                    w             └─────────────────┘

Security Layers

Node.js - Classifies intent, builds parameterized SELECT queries, calls NVIDIA LLM for natural responses
Java API - Validates SQL against strict regex patterns, executes only approved SELECT queries on EMPTAB
Database - Read-only credentials, FETCH FIRST 50 ROWS ONLY enforced

Video Demo

Db2.AI.Assistant.Demo.webm

Note: If inline video playback is not supported by your Markdown renderer, click here to open or download the WebM video directly.

Features

Natural Language Queries: Ask questions like "Show me employee John" or "What's the mobile for ID 1001"
System Health Monitoring: Real-time status of Node.js, NVIDIA API, Java DB2 API, JDBC, DDF, and Db2
Read-Only Access: Only SELECT on EMPTAB (EMPID, EMPNAME, EMPMOB) permitted
Context Awareness: Follow-up questions like "now" or "again" maintain conversation context
Security Guards: Blocks credential requests, mutation attempts, and out-of-scope queries
Dual Frontend: Modern React (Vite + TypeScript + Tailwind) and Streamlit Python UI

Project Structure

Ai_chatbot/
├── server.js                 # Node.js Express API (port 3001)
├── package.json              # Node.js dependencies
├── .env                      # Environment variables
├── images/
│   ├── Ai ChatBot Architecture - V1.png    # Architecture diagram
│   └── Db2 AI Assistant Demo.html          # Interactive demo
├── frontend/
│   ├── app.py               # Streamlit UI (port 8501)
│   └── requirements.txt     # Python dependencies
├── React_frontend/          # React + Vite + TypeScript UI
│   ├── src/
│   │   ├── components/      # Chat UI components
│   │   ├── hooks/           # Custom React hooks
│   │   ├── lib/             # API utilities
│   │   └── main.tsx         # Entry point
│   └── package.json
└── db2jdbc/
    ├── Db2RestServer.java   # Java HTTP server (port 8080)
    ├── Db2Test.java         # Connection test utility
    └── db2jcc4.jar          # Db2 JDBC driver

Prerequisites

Node.js 18+
Java 17+ (for Db2 JDBC driver)
Python 3.9+ (for Streamlit frontend)
Db2 for z/OS accessible via DDF (host:port/database)
NVIDIA API Key for LLM integration

Quick Start

1. Configure Environment

Create .env file in project root:

# NVIDIA LLM (required for natural language responses)
NVIDIA_API_KEY=your_nvidia_api_key
NVIDIA_URL=https://integrate.api.nvidia.com/v1/chat/completions
NVIDIA_MODEL=openai/gpt-oss-20b

# Java DB2 API
JAVA_DB2_API=http://127.0.0.1:8080/query
JAVA_HEALTH_API=http://127.0.0.1:8080/health
JAVA_DB2_STATUS_API=http://127.0.0.1:8080/db2/status

# Db2 Connection (used by Java API)
DB2_URL=jdbc:db2://your-mainframe-host:5025/DALLAS9
DB2_USER=IBMUSER
DB2_PASSWORD=your_db2_password
DB2_SERVER=DALLAS9

# Node.js API
PORT=3001

2. Start Java DB2 API

cd db2jdbc
# Compile
javac -cp ".:db2jcc4.jar" Db2RestServer.java
# Run
java -cp ".:db2jcc4.jar" Db2RestServer

Server starts on http://127.0.0.1:8080

3. Start Node.js API

npm install
npm start
# or: node server.js

Server starts on http://127.0.0.1:3001

4. Start Frontend (choose one)

React Frontend:

cd React_frontend
npm install
npm run dev

Opens at http://localhost:5173

Streamlit Frontend:

cd frontend
pip install -r requirements.txt
streamlit run app.py

Opens at http://localhost:8501

API Endpoints

Node.js API (Port 3001)

Method	Endpoint	Description
POST	`/chat`	Main chat endpoint
GET	`/health`	Health check
GET	`/system-status`	Full system status
GET	`/access`	Available tables/columns

Chat Request:

{
  "message": "Show me employee John",
  "history": [
    {"role": "user", "content": "previous question"},
    {"role": "assistant", "content": "previous answer"}
  ]
}

Chat Response:

{
  "success": true,
  "type": "DATA_QUERY",
  "answer": "Found 1 employee:\n\nEmployee ID: 1001\nName: JOHN DOE\nMobile: 555-0123",
  "data": [{"EMPID": "1001", "EMPNAME": "JOHN DOE", "EMPMOB": "555-0123"}],
  "query": {"operation": "SELECT", "table": "EMPTAB", "columns": ["EMPID", "EMPNAME", "EMPMOB"], "limit": 20}
}

Java DB2 API (Port 8080)

Method	Endpoint	Description
GET	`/health`	Basic service health
GET	`/db2/status`	Db2/JDBC/DDF health check
POST	`/query`	Execute validated SELECT query

Supported Query Types

Type	Examples
Greeting	"Hi", "Hello"
Goodbye	"Bye", "See you"
Data Query	"Show employee John", "What's ID 1001 mobile?", "List all employees"
System Status	"System status", "Is Db2 connected?", "Health check"
Access Info	"What tables can I access?", "Show schema"
Mainframe General	"How does the mainframe work?", "Explain DDF"

Security Features

Input Classification: Local + LLM-based intent detection
SQL Injection Prevention: Parameterized queries, regex validation on both Node.js and Java layers
Mutation Blocking: INSERT/UPDATE/DELETE/DROP/etc. rejected at both layers
Credential Protection: Detects and blocks requests for passwords, API keys, connection strings
Row Limits: Maximum 50 rows per query
Column Whitelist: Only EMPID, EMPNAME, EMPMOB accessible

Environment Variables

Node.js (.env)

Variable	Default	Description
`PORT`	3001	Node.js server port
`NVIDIA_API_KEY`	-	Required NVIDIA API key
`NVIDIA_URL`	NVIDIA integrate endpoint	LLM API URL
`NVIDIA_MODEL`	openai/gpt-oss-20b	Model to use
`JAVA_DB2_API`	http://127.0.0.1:8080/query	Java query endpoint
`JAVA_HEALTH_API`	http://127.0.0.1:8080/health	Java health endpoint
`JAVA_DB2_STATUS_API`	http://127.0.0.1:8080/db2/status	Java Db2 status endpoint
`DB2_SERVER`	DALLAS9	Db2 server name for display

Java (Environment)

Variable	Default	Description
`JAVA_API_HOST`	127.0.0.1	Java server bind address
`JAVA_API_PORT`	8080	Java server port
`DB2_URL`	jdbc:db2://192.168.200.150:5025/DALLAS9	JDBC connection URL
`DB2_USER`	IBMUSER	Db2 username
`DB2_PASSWORD`	-	Required Db2 password

Development

Run Tests

# Test Java DB2 connection
cd db2jdbc
javac -cp ".:db2jcc4.jar" Db2Test.java
java -cp ".:db2jcc4.jar" Db2Test

Build React for Production

cd React_frontend
npm run build

Lint/Typecheck React

cd React_frontend
npm run lint
npm run typecheck

Troubleshooting

Issue	Solution
"NVIDIA_API_KEY is not configured"	Add key to `.env`
"Java DB2 API is unreachable"	Ensure Java server running on port 8080
"DB2_PASSWORD environment variable is not set"	Set `DB2_PASSWORD` in environment
"Db2 query failed"	Check DDF connectivity, credentials, table existence
CORS errors	Node.js allows all origins (`*`) by default

License

MIT

Navigation Menu

Repository navigation

About

Repository files navigation

AI Chatbot for Mainframe Data

Architecture Overview

Architecture V1

Security Layers

Video Demo

Features

Project Structure

Prerequisites

Quick Start

1. Configure Environment

2. Start Java DB2 API

3. Start Node.js API

4. Start Frontend (choose one)

API Endpoints

Node.js API (Port 3001)

Java DB2 API (Port 8080)

Supported Query Types

Security Features

Environment Variables

Node.js (.env)

Java (Environment)

Development

Run Tests

Build React for Production

Lint/Typecheck React

Troubleshooting

License

About

Releases

Packages

Contributors

Languages

Name		Name	Last commit message	Last commit date
React_frontend		React_frontend
db2jdbc		db2jdbc
frontend		frontend
images		images
.gitignore		.gitignore
README.md		README.md
package-lock.json		package-lock.json
package.json		package.json
server.js		server.js
server.log		server.log
test_server.sh		test_server.sh
test_server2.sh		test_server2.sh

Latest commit

History

Folders and files

Repository files navigation

AI Chatbot for Mainframe Data

Architecture Overview

Architecture V1

Security Layers

Video Demo

Features

Project Structure

Prerequisites

Quick Start

1. Configure Environment

2. Start Java DB2 API

3. Start Node.js API

4. Start Frontend (choose one)

API Endpoints

Node.js API (Port 3001)

Java DB2 API (Port 8080)

Supported Query Types

Security Features

Environment Variables

Node.js (.env)

Java (Environment)

Development

Run Tests

Build React for Production

Lint/Typecheck React

Troubleshooting

License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages