AidjuDigital
AI-powered retail business management platform. Query data, generate documents, run reports, and automate workflows — all through natural language.
160+
Functions
26
Document Types
8
Dashboards
13
Categories
Three AI Engines
Working together to answer any business question from every angle
Vanna AI
Text-to-SQL Engine
Converts natural language questions into precise SQL queries. Trained on the retail database schema, it generates accurate queries, executes them, and returns results with explanations.
EdgeQuake
Knowledge Graph Engine
Builds and searches a knowledge graph of business entities and relationships using Apache AGE. Maps connections between customers, products, suppliers, and transactions.
PageIndex
Document Retrieval Engine
A reasoning-based RAG system that builds hierarchical tree indexes from documents. Uses LLM reasoning to navigate document structure like a human expert — finding the most relevant sections with traceable evidence.
How a Question Gets Answered
User Asks a Question
"What were the top 5 selling products last month and what does the inventory report say about their stock levels?"
Vanna Generates SQL
Converts the question to a precise SQL query, executes it against PostgreSQL, and returns structured data with product names, revenue, quantities, and rankings.
EdgeQuake Searches the Graph
Finds related entities in the knowledge graph — connected suppliers, customer segments, product categories — providing contextual understanding.
PageIndex Searches Documents
Reasons over the hierarchical index of uploaded documents, navigates to the most relevant sections, and extracts evidence with specific citations.
AI Synthesizes the Answer
Gemini LLM combines data from all three sources — numbers from Vanna, context from EdgeQuake, evidence from PageIndex — into a comprehensive, data-driven response with auto-generated charts.
Platform Features
Dashboard
Executive KPI overview with performance metrics and activity feed
AI Chat
Ask any question — powered by Vanna, EdgeQuake, and PageIndex working together
Data Upload
Upload Excel, CSV, PDF, Word files for AI-powered analysis and ingestion
Knowledge Graph
Interactive force graph visualization of entity relationships
Sales
22 sales reports — daily, weekly, monthly, by category, product, segment
Customers
13 customer reports — purchase history, CLV, segmentation, acquisition
Finance
13 finance reports — P&L, cash flow, SST, commissions, budgets
Documents
26 document types — invoices, POs, receipts, statements, and more
Analytics
8 role-specific dashboards with live KPIs and charts
AI Chat
Your primary interface to the entire system
Natural Language Queries
Ask questions in plain English, Bahasa Melayu, or Chinese. The AI translates your intent into database queries, graph lookups, or document actions.
3-Layer Retrieval
Combines SQL generation (Vanna), vector semantic search (PageIndex), and knowledge graph traversal (EdgeQuake) for comprehensive answers.
Auto Chart Generation
Responses with numerical data automatically generate interactive charts — bar, line, area, scatter, pie, radar, treemap, or heatmap.
Document Generation
Say "Create a quotation for 4x RTX 4090 to Ali" and the AI composes a branded PDF with product lookups, pricing, and tax calculations.
Email Integration
Say "Email that invoice to Ali" — the AI composes a professional email, attaches the PDF, and sends it via your configured SMTP.
File Analysis
Attach CSV, Excel, PDF, or Word files in chat. The AI analyzes the content and answers questions about your uploaded data.
Supported Chart Types
Multilingual Support
English, Bahasa Melayu, Simplified Chinese
Document Generation
26 document types, generated from chat or function pages
Customer-Facing (18)
- Sales Invoice
- Proforma Invoice
- Credit Note
- Debit Note
- Payment Receipt
- Quotation Letter
- Delivery Order
- Packing List
- Warranty Card
- Service Job Sheet
- Service Invoice
- RMA Acknowledgment
- Statement of Account
- Loyalty Points Statement
- Product Price List
- Gift Receipt
- Collection Notice
- Refund Voucher
Internal (8)
- Purchase Order
- Goods Received Note
- Stock Transfer Note
- Stock Count Sheet
- Expense Claim Form
- Commission Slip
- Payroll Summary
- Supplier Payment Voucher
Smart Document Features
Function Categories
160+ functions across reports, documents, dashboards, and tools
Sales
RSAL-01 to RSAL-22
Customers
RCUS-01 to RCUS-13
Products
RINV-10, 11, 14, 15
Inventory
RINV-01 to RINV-09, 12, 13
Purchasing
RSUP-03, RSUP-05
Suppliers
RSUP-01, 02, 04, 06-08
Finance
RFIN-01 to RFIN-13
Staff
RSTF-01 to RSTF-07
Service & Warranty
RSVC-01 to RSVC-11
Loyalty
RLOY-01 to RLOY-07
Documents
DOC-01 to DOC-18, IDOC-01 to IDOC-08
Tools
TOOL-01 to TOOL-18
Dashboards
DASH-01 to DASH-08
Analytics
RLOY-01 to RLOY-07
Dashboards
8 role-specific dashboards with live KPIs and charts
Executive
High-level KPIs and company overview
Store Manager
Operational metrics and daily performance
Inventory
Stock health, movement, and alerts
Customer
Behavior, segmentation, and loyalty
Financial
P&L, cash flow, and health indicators
Service Center
Job pipeline, turnaround, and technician load
Supplier
Performance scorecard and spend analytics
Year-End
Annual review with full-year KPIs
Tools
18 utility functions for lookups, calculations, and operations
Lookups
- Customer Lookup (name, phone, email)
- Product Lookup (name, SKU, barcode)
- Order Lookup (order number)
- Price Check
- Stock Check
Calculators
- Discount Calculator
- Currency Converter (MYR, USD, SGD, EUR, GBP)
- Trade-In Calculator
- PC Compatibility Checker
- Data Validation Checker
Operations
- Quotation to Invoice Conversion
- Bulk Price Update
- Auto Reorder (PO generation)
- Delivery Route Planner
- Barcode Label Generator
- Document Number Sequencer
- Bundle Creator
- Customer Duplicate Checker
Data Upload & Ingestion
Upload your business data for AI-powered analysis
Supported Formats
4-Stage Pipeline
- 1 Parse — file validation & format conversion
- 2 Embed — AI vectorization for semantic search
- 3 Graph — entity extraction & knowledge graph
- 4 Upsert — database insert/update
Technology Stack
Built for reliability, speed, and intelligent retrieval
Frontend
- Next.js 15
- React 19
- Tailwind CSS
- Recharts
- Streamdown
Backend
- Node.js
- Drizzle ORM
- Flask (Python)
- Gunicorn
AI / LLM
- Gemini 2.5 Flash
- Vanna AI
- EdgeQuake
- PageIndex
- LiteLLM
Infrastructure
- PostgreSQL + Apache AGE
- pgvector
- Redis
- Docker
- SSH Tunnels
Dual AI Backend
EdgeQuake for knowledge-rich graph queries, Google Gemini for SQL generation, intent detection, and chart composition.
PostgreSQL + pgvector
Relational data with vector embedding support for semantic similarity search across all uploaded documents.
Knowledge Graph
Apache AGE graph database for entity relationships — customers, products, suppliers, and their connections.
Real-Time Streaming
Server-Sent Events for token-by-token AI responses with inline chart rendering and tool result cards.
Security
Read-only database access for AI queries, input sanitization, parameterized SQL, and concurrency semaphores.
Feature Flags
Per-function kill switches with Redis caching. Enable or disable any of the 160+ functions without code changes.