🚀 DataPilot Backend - AI-Powered Salesforce Platform

🌟 THE AI REVOLUTION FOR SALESFORCE BACKEND 🌟

Transform Salesforce operations with AI-powered backend intelligence, autonomous agents, and enterprise-grade security.

📋 Table of Contents

ARCHITECTURE OVERVIEW

Core Technologies

FastAPI: High-performance async web framework
MongoDB: Document database for data persistence
LangGraph: AI workflow orchestration
WebSocket: Real-time AI streaming

AI Integration

Custom ReAct Agent: Autonomous AI reasoning and acting
LangGraph Workflows: Stateful AI conversation management
Real-time Streaming: Live AI responses via WebSocket
Natural Language Processing: Convert queries to SOQL

Architecture Overview

High-Level Architecture

graph LR
    A[Frontend<br/>React + TypeScript] --> B[Backend API<br/>FastAPI]
    B --> C[WebSocket<br/>Real-time Streaming]
    C --> D[AI Engine<br/>LangGraph]
    B --> E[Database<br/>MongoDB]
    D --> F[Salesforce API]
    D --> G[LLM Providers<br/>OpenAI/Groq/Ollama]
    C --> A

AI Engine Architecture

graph TB
    A[Natural Language Input] --> B[LangGraph Workflow State]
    B --> C[Agent Node]
    C --> D[LLM with System Prompt]
    D --> E[AI Response with Tool Calls]
    E --> F[Should Continue Node]
    F --> G{Has Tool Calls?}
    G -->|Yes| H[Tool Node]
    G -->|No| I[Summary Node]
    H --> J[Execute Salesforce Tools]
    J --> K[search_for_sobjects]
    J --> L[get_sobject_metadata]
    J --> M[get_sobject_relationships]
    J --> N[execute_soql_query]
    J --> O[get_field_details]
    K --> P[Salesforce API]
    L --> P
    M --> P
    N --> P
    O --> P
    P --> Q[Tool Results]
    Q --> R[Redact for LLM]
    R --> S[Tool Messages]
    S --> C
    I --> T[Structured JSON Response]
    T --> U[Real-time Streaming]
    U --> V[Frontend Display]

QUICK START

Prerequisites

Python 3.11+
uv package manager
MongoDB (local or cloud)
Langfuse (optional, for AI agent observability and monitoring)

Installation & Setup

# 1. Install uv (if not already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh

# 2. Clone and navigate to backend
cd backend

# 3. Install dependencies
make install-dev

# 4. Configure environment
cp env.example .env
# Edit .env with your settings

# 5. Start the backend
make run

Alternative Startup Methods

# Method 1: Using Makefile (Recommended)
make run                    # Development server with auto-reload
make run-prod              # Production server
make dev                   # Alias for run

# Method 2: Using start.sh script
./start.sh                 # Automated setup and startup

# Method 3: Direct uv command
uv run python -m uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

API ENDPOINTS OVERVIEW

Core Services

Service	Endpoint	Description	Model
Health	`GET /api/v1/health`	System health and status	`HealthResponse`
Salesforce	`/api/v1/salesforce/*`	SOQL queries, SObjects, Apex execution	`QueryResponse`, `SObjectMetadata`
Connections	`/api/v1/connections/*`	Secure connection management	`ConnectionResponse`, `ConnectionCreate`
AI Agents	`/api/v1/ai-agents/*`	AI conversation and workflows	`AIQueryRequest`, `AIQueryResponse`
Master Key	`/api/v1/master-key/*`	Security and encryption	`MasterKeyResponse`, `MasterKeyCreate`

AI & Intelligence

Endpoint	Method	Capability	Model
`/ai-agents/{conversation_uuid}`	POST	Natural language to SOQL conversion	`AIQueryRequest`, `AIQueryResponse`
`/ai-agents/{conversation_uuid}/stream`	WebSocket	Real-time AI streaming	`StreamingResponse`
`/ai-agents/conversations`	GET/POST	Conversation management	`ConversationResponse`, `ConversationCreate`
`/ai-agents/conversations/{uuid}`	GET/PUT/DELETE	Individual conversation operations	`ConversationResponse`, `ConversationUpdate`

Security & Authentication

Endpoint	Method	Capability	Model
`/master-key/`	GET/POST/DELETE	Master key management	`MasterKeyResponse`, `MasterKeyCreate`
`/connections/`	GET/POST	Connection lifecycle	`ConnectionResponse`, `ConnectionCreate`
`/connections/{uuid}`	GET/PUT/DELETE	Individual connection operations	`ConnectionResponse`, `ConnectionUpdate`
`/auth-providers/`	GET/POST	Authentication providers	`AuthProviderResponse`, `AuthProviderCreate`

Data Operations

Endpoint	Method	Capability	Model
`/salesforce/queries/execute`	POST	SOQL query execution	`QueryRequest`, `QueryResponse`
`/salesforce/sobjects/list`	GET	SObject discovery	`SObjectListResponse`
`/salesforce/sobjects/describe/{name}`	GET	SObject metadata	`SObjectMetadata`
`/salesforce/apex/execute-anonymous`	POST	Apex code execution	`ApexExecuteRequest`, `ApexExecuteResponse`
`/saved-queries/`	GET/POST	Query management	`SavedQueryResponse`, `SavedQueryCreate`
`/saved-apex/`	GET/POST	Apex code management	`SavedApexResponse`, `SavedApexCreate`

Internationalization

Endpoint	Method	Capability	Model
`/i18n/translations/{lang}`	GET	Get translations	`TranslationResponse`
`/i18n/languages`	GET	Available languages	`LanguageResponse`
`/i18n/languages`	POST	Add new language	`LanguageCreate`, `LanguageResponse`

Monitoring & Logging

Endpoint	Method	Capability	Model
`/logging/`	GET/POST	Log management	`LogResponse`, `LogCreate`
`/settings/`	GET/PUT	Application settings	`SettingsResponse`, `SettingsUpdate`
`/sobjects/cache`	GET	Cache statistics	`CacheStatsResponse`

DEVELOPMENT COMMANDS

Installation & Dependencies

make install          # Production dependencies
make install-dev      # All dependencies (including dev)
make sync             # Sync from lock file
make update           # Update all dependencies

MongoDB Setup

# Start MongoDB
mongod --dbpath /path/to/data

# Or use Docker
docker run -d -p 27017:27017 --name mongodb mongo:latest

AI CAPABILITIES

Autonomous AI Agent

ReAct Pattern: Reasoning + Acting for complex operations
Natural Language Processing: Convert queries to SOQL
Real-time Streaming: Live AI responses via WebSocket
Context Awareness: Maintains conversation state
Tool Integration: 5 specialized Salesforce tools

AI Tools

search_for_sobjects: SObject discovery and search
get_sobject_metadata: Detailed SObject information
get_sobject_relationships: Relationship mapping
execute_soql_query: Query execution and optimization
get_field_details: Field-level metadata and validation

MONITORING & OBSERVABILITY

Health Monitoring

System Health: /api/v1/health endpoint
Service Status: Real-time availability monitoring
Error Tracking: Comprehensive error logging

Logging & Debugging

Structured Logging: JSON-formatted logs
Error Reporting: Detailed error information

API DOCUMENTATION

Interactive Documentation

Swagger UI: http://localhost:8000/docs
ReDoc: http://localhost:8000/redoc
OpenAPI Schema: http://localhost:8000/openapi.json

API Examples

# Health check
curl http://localhost:8000/api/v1/health

# List SObjects
curl http://localhost:8000/api/v1/salesforce/sobjects/list

# Execute SOQL
curl -X POST http://localhost:8000/api/v1/salesforce/queries/execute \
  -H "Content-Type: application/json" \
  -d '{"query": "SELECT Id, Name FROM Account LIMIT 5"}'

LICENSE

This project is licensed under the DataPilot License - see the LICENSE file for details.

Commercial and Enterprise Use: Requires prior written permission. Contact: https://www.linkedin.com/in/bassem-elsodany/

🚀 GET STARTED NOW

Ready to revolutionize your Salesforce operations with AI?

# Quick start
make install-dev && make run

🌐 Backend running at: http://localhost:8000
📚 API Documentation: http://localhost:8000/docs

🤖 THE AI REVOLUTION FOR SALESFORCE BACKEND IS HERE 🤖

Transform your Salesforce operations with AI-powered intelligence, autonomous agents, and enterprise-grade security.

FilesExpand file tree

README.md

Latest commit

History