🗂️ TaskFlow- Task Management Application

🧑‍💻 Full‑Stack Development — Next.js ⚛️ | .NET Web API 🛠️ | SQL Server 🗄️ | Python NER 🤖

A production‑grade Task Management system featuring real‑time task transitions, instant updates across Todo, In‑Progress, and Completed states, secure authentication, and a clean full‑stack architecture. Now enhanced with intelligent task insights powered by:

• Python‑based Named Entity Recognition (NER) for automatic due‑date detection and priority suggestions based on keywords in the title (e.g., “urgent”, “tomorrow”).
• ML.NET machine learning model that predicts task priority directly from the task’s title and description, automatically assigning a priority when none is selected. Together, these AI features deliver a smarter, more intuitive task creation and editing experience.

🖥️ Frontend — Next.js

• Next.js 14 (App Router)

• TypeScript

• Tailwind CSS

• shadcn/ui

• Client-side routing

🔐 Backend — .NET Web API

• .NET 8 Web API

• Entity Framework Core

• JWT Authentication

• HttpOnly Cookies

• SQL Server

• DTO validation

• Role-based authorization

AI Integration (Two Approaches)

My application uses two different AI systems, each serving a unique purpose.
Both approaches enhance the task creation and editing experience in different ways.

---

🤖 Python NER Service (Natural Language Processing)

• Full‑stack implementation using .NET (API) and Next.js (UI)
• Python‑based Named Entity Recognition (NER).
• Extracts dates, keywords, and priority‑related hints from the task description.
• Shows toast notifications with suggestions and auto‑fills the recommended values into the form.
• User can still modify the suggestions before saving.
• Integrated into both Create Task and Edit Task flows.
• Communicates with the .NET backend to provide AI‑generated metadata.

How it works

• Runs locally using a Python script.
• Analyzes the task title to detect meaningful entities.
• Helps with automation, tagging, and metadata extraction.

Example

User enters:

“urgent users report by next Monday”

Python NER extracts:

• Date: next Monday
• Priority: High (because of the keyword “urgent”)

The system then:

• Automatically fills the due date as next Monday
• Automatically sets priority = High
• Still allows the user to change these values before saving

---

⚡ ML.NET Priority Prediction (Machine Learning)

• Fully implemented inside .NET using ML.NET.
• Uses the own labeled task data from the database.
• TaskTrainer trains the model and saves it.
• Predicts task priority (1 = Low, 2 = Medium, 3 = High) based on title + description.
• Automatically assigns a priority when the user leaves it blank during task creation.
• Integrated with Swagger for testing and debugging.

How it works

• The model is trained using real tasks from the database.
• PredictPriorityAsync loads the model and predicts priority in real time.
• Used during task creation when no priority is selected.

Example

Title:

“Fix production bug”

Description:

“API returning 500 errors”

ML.NET prediction:
→ High Priority

Frontend - testing

• vitest

• MSW(Mock Service Worker)

• @testing-library/jest-dom

📸 Screenshots

Dashboard

Login

Register

Forgot Password

View Task

Create Task

Edit Task

Delete Task

Mobile Responsive Dashboard

Swagger API

Database Schema

---

🤖 Local python NER Integration

AI Suggested Priority (Create Task)

AI Suggested Priority (Edit Task)

Swagger – AI Output

Python Named Entity Recognition (NER) Integration (Local)

ML.NET Priority Prediction

ML Trained Data (Swagger)

AI Suggested Priority (Swagger)

ML Dotnet Suggested Priority based on user's title and description (Create Task)

📦 Setup

🚀 Frontend(Nextjs)

git clone https://github.com/vbudithi/task-management-app.git
cd task-management-app/frontend

npm install
npm run dev

🧩Backend(.NET API)

git clone https://github.com/vbudithi/task-management-app.git
cd backend/TaskManagement.API

dotnet restore
dotnet run

🤖Python NER Server

cd ai-ner-server
python ner_server.py