CognifyAI

AI-Powered Code Optimization & Analysis Platform

Analyze, secure, optimize, and document code using a multi-stage AI pipeline.

---

Overview

CognifyAI is a full-stack AI-assisted code analysis platform designed to help developers improve code quality through a structured analysis pipeline.

Users can submit source code and receive:

• Syntax validation
• Bug detection
• Security vulnerability analysis
• Complexity estimation
• Code optimization suggestions
• Automated docstring generation
• Overall quality scoring

The platform combines deterministic static analysis with optional LLM-powered enrichment, providing fast and reliable results while remaining cost-efficient.

---

Features

Multi-Stage Analysis Pipeline

Every code submission passes through a strictly sequential pipeline:

graph TD
    SV["Syntax Validation"] --> BD["Bug Detection"]
    BD --> SS["Security Scanning"]
    SS --> CA["Complexity Analysis"]
    CA --> CO["Code Optimization"]
    CO --> DG["Docstring Generation"]

    %% Styling for visual clarity
    style SV fill:#e1f5fe,stroke:#0288d1,stroke-width:2px,color:#000
    style DG fill:#e8f5e9,stroke:#388e3c,stroke-width:2px,color:#000

Loading

Analysis Capabilities

• AST-based static code analysis
• Bug detection with line-level explanations
• Security vulnerability scanning
• Time complexity estimation
• Space complexity estimation
• Code optimization suggestions
• Automatic documentation generation
• Quality scoring engine

Interactive Frontend

• Monaco Editor
• Live pipeline progress tracking
• Analysis dashboard
• Historical analysis records
• Side-by-side diff viewer
• Optimized code comparison

AI-Powered Enhancements

Supports two execution modes:

Deterministic Mode

• No API key required
• Fully local analysis
• Fast execution
• Rule-based heuristics

LLM Enrichment Mode

• LangChain integration
• Mistral AI support
• Enhanced explanations
• Smarter optimization recommendations
• More contextual code reviews

---

Architecture

graph TD
    FE["Frontend (React + TypeScript)"] --> BE["FastAPI Backend"]
    
    subgraph Core_Services ["Core Core Engine Layer"]
        BE --> PL["Pipeline Block"]
        BE --> DB_Layer["Database Interface"]
        BE --> LLM["LLM Layer (LangChain / Mistral)"]
    end
    
    DB_Layer --> DB[("PostgreSQL Database")]

    %% Styling
    style FE fill:#e1f5fe,stroke:#0288d1,stroke-width:2px,color:#000
    style BE fill:#e8f5e9,stroke:#388e3c,stroke-width:2px,color:#000
    style DB fill:#ede7f6,stroke:#5e35b1,stroke-width:2px,color:#000

Loading

---

Technology Stack

Backend

• Python 3.11
• FastAPI
• SQLAlchemy 2.0
• PostgreSQL
• Pydantic v2

Frontend

• React 18
• TypeScript
• Vite
• Monaco Editor

AI Layer

• AST Analysis
• LangChain
• Mistral AI

DevOps

• Docker
• Docker Compose
• Nginx

Testing

• Pytest

---

Project Structure

CognifyAI/
├── backend/
│   ├── app/
│   ├── tests/
│   ├── Dockerfile
│   └── requirements.txt
│
├── frontend/
│   ├── src/
│   ├── Dockerfile
│   └── package.json
│
├── docs/
│   ├── api_design.md
│   ├── architecture.md
│   └── prompts.md
│
├── screenshot/
│   ├── home.jpeg
│   ├── result.jpeg
│   └── history.jpeg
│
├── docker-compose.yml
└── README.md

---

Run With Docker

Clone Repository

git clone [https://github.com/maroofiums/CognifyAI.git](https://github.com/maroofiums/CognifyAI.git)
cd CognifyAI

Start Application

docker-compose up --build

Available Services

Service	URL
Frontend	http://localhost:5173
Backend API	http://localhost:8000
Swagger Docs	http://localhost:8000/docs
PostgreSQL	localhost:5432

The frontend automatically communicates with the backend through Nginx proxy configuration.

Stop Services

docker-compose down

Remove database volume:

docker-compose down -v

---

Enable LLM Layer

By default:

USE_LLM=false

To enable LangChain + Mistral enrichment:

environment:
  USE_LLM: "true"
  MISTRAL_API_KEY: "your-api-key"
  LLM_MODEL: "mistral-small-latest"

---

Local Development

Backend Setup

cd backend

python -m venv .venv

source .venv/bin/activate
# Windows
.venv\Scripts\activate

pip install -r requirements.txt

uvicorn app.main:app --reload --port 8000

Backend:

http://localhost:8000

Swagger Documentation:

http://localhost:8000/docs

---

Frontend Setup

cd frontend

npm install

npm run dev

Frontend:

http://localhost:5173

---

PostgreSQL Configuration

For local PostgreSQL:

DATABASE_URL=postgresql://cognify:cognify@localhost:5432/cognifydb

Run PostgreSQL:

docker run -d \
  --name cognify-db \
  -e POSTGRES_USER=cognify \
  -e POSTGRES_PASSWORD=cognify \
  -e POSTGRES_DB=cognifydb \
  -p 5432:5432 \
  postgres:16-alpine

---

Using CognifyAI

1. Open the Home page and paste source code into the Monaco editor.
2. Select the programming language.
3. Click Analyze Code.
4. Watch live pipeline progress updates.
5. Review:

• Quality score
• Security findings
• Bug reports
• Complexity analysis
• Generated documentation
• Optimized code suggestions

6. Browse historical analyses from the History page.

---

API Example

Request

{
  "language": "python",
  "code": "def add(a,b): return a+b"
}

Response

{
  "bugs": [],
  "security_issues": [],
  "complexity": {
    "time": "O(1)",
    "space": "O(1)"
  },
  "optimized_code": "def add(a: int, b: int) -> int:\n    return a + b",
  "docstring": "Returns the sum of two integers.",
  "score": {
    "correctness": 95,
    "readability": 90,
    "security": 100,
    "performance": 95,
    "documentation": 85,
    "overall": 93
  }
}

---

JSON Output Contract

{
  "bugs": [],
  "security_issues": [],
  "complexity": {
    "time": "O(n)",
    "space": "O(1)"
  },
  "optimized_code": "",
  "docstring": "",
  "score": {
    "correctness": 0,
    "readability": 0,
    "security": 0,
    "performance": 0,
    "documentation": 0,
    "overall": 0
  }
}

---

Testing

Run the complete backend test suite:

cd backend

pytest -q

Coverage includes:

• API tests
• Pipeline tests
• Service tests
• Security checks
• Syntax validation
• Score calculation

---

Documentation

Additional documentation can be found inside the docs directory.

File	Purpose
architecture.md	System architecture
api_design.md	API documentation
prompts.md	LLM prompt templates

---

Future Roadmap

Version 2

• Redis integration
• Celery background workers
• WebSocket status updates
• Authentication & authorization
• Multi-language support
• Team workspaces

Version 3

• LangGraph orchestration
• RAG-powered recommendations
• GitHub repository analysis
• Pull request reviews
• CI/CD integration
• AI code review reports

---

Why CognifyAI?

Most code review tools focus on one aspect of software quality. CognifyAI combines Static Analysis, Security Review, Optimization, Documentation Generation, and AI-Assisted Insights into a single developer workflow.

The project demonstrates production-grade software engineering concepts including:

• Clean Architecture
• Repository Pattern
• Service Layer Pattern
• REST API Design
• Containerized Deployment
• AI Workflow Pipelines
• Frontend-Backend Integration

---

License

MIT License

Feel free to use, modify, and contribute.