Academic Email Automation

Smart Email Management System for Students using FastAPI, UiPath RPA, and LLMs

Overview

This project demonstrates an intelligent email-automation platform designed to help students manage the overwhelming volume of placement-related communication. Students at VIT receive dozens of emails every week containing recruitment announcements, deadlines, shortlists, and general academic notifications. Monitoring these manually is time-consuming and often leads to missed opportunities.

This system automates the entire workflow:

• Extracting emails
• Classifying important messages
• Performing eligibility checks
• Reading shortlist/selection attachments
• Summarising long emails using an LLM
• Generating ICS calendar invites via UiPath
• Displaying all results through a clean web interface

The solution integrates UiPath RPA for inbox automation and FastAPI + SQLite + HTML/JS for backend and frontend logic, creating a fully functional end-to-end academic email assistant.

---

Table of Contents

1. Project Motivation
2. Key Features
3. Architecture
4. Tech Stack
5. System Workflow
6. Installation
7. Running the System
8. UiPath Workflow Integration
9. Email Classification Logic
10. Limitations & Future Work
11. Contributors

---

Project Motivation

Students are exposed to an overwhelming volume of daily emails, and manually identifying critical placement-related communication is inefficient and error-prone. Important details such as eligibility criteria, deadlines, interview schedules, and shortlists are often missed.

This project solves that problem by creating an automated system that transforms raw inbox data into structured, actionable insights that students can trust.

---

Key Features

1. Automated Email Extraction (UiPath)

• Reads unread inbox emails
• Downloads and scans attachments
• Sends extracted data to backend for analysis

2. Intelligent Email Classification

Emails are automatically categorized into:

1. New Company Announcement
2. Shortlist / Round Progression
3. Final Selection
4. General Academic Notices

Classification uses both keyword logic and LLM-based semantic understanding.

3. Eligibility Checker

• Extracts CGPA, branch, 10th/12th marks from the user profile (SQLite)
• Compares with company criteria
• Displays “Eligible” / “Not Eligible” instantly

4. Excel Shortlist Processing

• Parses Excel sheets automatically
• Checks if student's name or registration number appears
• Works even when names are inside attachments

5. LLM-Based Email Summaries

Long announcements are summarized intelligently to quickly surface key information.

6. Calendar Automation (ICS Generation)

Via UiPath:

• Automatically identifies dates from emails
• Generates ICS files
• Sends them to student mailbox for one-click calendar import

7. Web Dashboard

Includes:

• Signup & login
• User profile view
• Dashboard showing categorized emails
• Embedded summaries, calendar buttons

---

Architecture

High-Level Architecture Diagram

                ┌──────────────┐
                │ Student Inbox│
                └───────┬──────┘
                        │ (RPA reads emails)
                  ┌─────▼──────┐
                  │   UiPath   │
                  │  Workflows │
                  └─────┬──────┘
                        │ Extracted data (JSON + files)
                  ┌─────▼──────────┐
                  │   FastAPI API  │
                  └─────┬──────────┘
                        │
      ┌─────────────────▼──────────────────┐
      │ Classification | Eligibility | LLM │
      └─────────────────┬──────────────────┘
                        │
               ┌────────▼─────────┐
               │  SQLite Database │
               └────────┬─────────┘
                        │
              ┌─────────▼──────────┐
              │   Web Frontend     │
              │  (HTML/CSS/JS)     │
              └────────────────────┘

Each component communicates seamlessly, forming a hybrid RPA + AI-driven academic assistant.

---

Tech Stack

Backend

• Python
• FastAPI
• SQLite (users.db)
• SQLAlchemy ORM

Frontend

• HTML
• CSS
• JavaScript

Automation (RPA)

• UiPath Studio
• Email Activities
• Excel Processing
• File Handling
• ICS Event Creation

AI / NLP

• LLM-based summarization
• Semantic extraction
• Deadline recognition

System Workflow

1. UiPath Workflow

• Fetch unread emails
• Extract content & attachments
• Send POST request to FastAPI

2. FastAPI Processing

• Parse email
• Detect category
• Run eligibility logic
• Scan Excel files if needed
• Generate summary using NLP
• Store results into SQLite DB

3. User Dashboard

• Students view categorized alerts
• Can add deadlines to calendar
• Can view summaries & decisions instantly

---

Installation

Step 1: Clone Repository

git clone https://github.com/allen-reji/Academic-Email-Automation
cd Academic-Email-Automation

Step 2: Create Virtual Env

python -m venv venv
source venv/bin/activate         (Linux/Mac)
venv\Scripts\activate            (Windows)

Step 3: Install Dependencies

pip install -r requirements.txt

Step 4: Initialize Database

The repository already includes users.db. If you wish to recreate it:

python
>>> from app.database import Base, engine
>>> Base.metadata.create_all(engine)
>>> exit()

---

Running the System

Start Backend Server

uvicorn app.main:app --reload --port 5000

Access Frontend

Open:

http://localhost:5000

HTML files are located under:

app/static/

---

UiPath Workflow Integration

Two main workflows are used:

1. Email Extraction Bot

• Reads student inbox
• Downloads Excel/PDF attachments
• Sends content to FastAPI using HTTP Request activity

2. ICS Calendar Bot

• Creates ICS file
• Sends it to student mailbox
• Triggered via backend request

The workflows correspond to .nupkg files included in the repo.

---

Email Classification Logic

Type 1 — Company Announcements

• Extracts CGPA, backlog criteria
• Matches with user data
• Returns Eligible / Not Eligible

Type 2 — Shortlist Notifications

• Reads Excel attachments
• Checks for student registration number

Type 3 — Final Selection

• Detects student name in selection lists

Type 4 — General Academic Notices

• LLM generates readable summary

---

Limitations & Future Work

Current Limitations

• Not yet deployed for large-scale multi-user usage
• Limited to VIT-style email formats
• Backend uses SQLite, not cloud database
• LLM integration is local; cloud models can be added later

Planned Improvements

• Mobile app with push notifications
• Multi-user authentication system
• Cloud deployment with PostgreSQL
• Advanced deadline extraction with NER
• Batch processing of emails

---

Contributors

Name	Email ID
Nidhish Balasubramanya	nidhishbalasubramanya@gmail.com
A Nathania Rachael	nathaniarachael@gmail.com
Allen Reji	allenreji@gmail.com
Jacob Cherian M	jakecherian10@gmail.com