Educational Institution Info Chatbot App
Budget: ₹600 – ₹1,500 INR
We developing a app to help students get information easily about their colleges and institutions.
Complete Application Flow
1. App Launch
The app opens with a Splash Screen (shows logo or animation).
After a few seconds, transitions to the Main Screen.
---
2. Main Screen UI
Contains a Top Navigation Bar with:
History – Shows all URLs previously queried by the user.
Settings – Includes preferences (e.g., notifications, appearance).
About App – Displays app details, version, and developer info.
Main content includes:
Input field to enter a website URL (educational institutions only).
Button to submit the URL.
Field to ask a question (enabled only when data is ready).
---
3. URL Entry & Validation
User enters a website URL.
The app checks if the URL belongs to an educational institution.
Valid examples: .edu, .ac.in, .edu.au, etc.
If invalid, show error message:
“Please enter a valid educational institution’s website.”
If valid, proceed to backend check.
---
4. Backend Check
The app sends the URL to the backend.
Backend checks if the website’s data already exists in the vector database.
---
5. If Data Exists
The app enables the question input field.
User can now enter a natural-language query.
The backend uses an AI model (e.g., GPT or Mistral) to generate and return an answer.
The answer is displayed in the chat interface.
---
6. If Data Does Not Exist
The backend starts scraping the website:
General text content
Contact information (emails, phone numbers, addresses)
Documents (PDFs, DOCs, etc.)
Scraping Progress UI
While scraping:
The app shows a progress screen with:
A loading spinner or progress bar
A message:
“Gathering information from the website. Please wait...”
Optional step updates:
“Scanning pages...”
“Extracting contact details...”
“Analyzing documents...”
“Finalizing...”
When scraping is complete:
A success message is shown:
“Data ready! You can now ask questions.”
The question input field is activated.
7. Asking a Question
User types a question.
Backend:
Searches the website's vectorized data.
Uses an AI model to answer.
The app displays the AI-generated human-readable response.
8. Data Maintenance
Re-scraping:
Each website is automatically re-scraped every 2 days to keep content fresh.
Auto-cleanup:
If a website’s data is not used for 30 days, it is deleted from the vector database.
9. History
The app maintains a list of all URLs entered by the user.
Accessible via the History tab in the top navigation bar.
Project Description
We are developing an Android chatbot application that answers user queries about any educational institute based on its website. The app should allow users to enter the URL of an institute’s website and then ask natural-language questions.
The backend will check whether the website’s data is already stored. If not, it will scrape the website for:
General text content
Contact information (emails, phone numbers, addresses)
Documents (PDFs, DOCs, etc.) with their contents extracted
The scraped data will be embedded and stored in a vector database. An AI language model (OpenAI, Mistral, or similar) will then use this data to generate human-like answers to user questions. If the data is already available, the user can immediately query it.
Additionally, the system must:
Maintain a list of URLs queried by each user
Automatically re-scrape and update the data for each URL every alternate day
Be crash-proof with robust error handling
The backend should be hosted on any free-tier platform (e.g., Render, Railway, Vercel).
Complete Application Flow
1. App Launch
The app opens with a Splash Screen (shows logo or animation).
After a few seconds, transitions to the Main Screen.
---
2. Main Screen UI
Contains a Top Navigation Bar with:
History – Shows all URLs previously queried by the user.
Settings – Includes preferences (e.g., notifications, appearance).
About App – Displays app details, version, and developer info.
Main content includes:
Input field to enter a website URL (educational institutions only).
Button to submit the URL.
Field to ask a question (enabled only when data is ready).
---
3. URL Entry & Validation
User enters a website URL.
The app checks if the URL belongs to an educational institution.
Valid examples: .edu, .ac.in, .edu.au, etc.
If invalid, show error message:
“Please enter a valid educational institution’s website.”
If valid, proceed to backend check.
---
4. Backend Check
The app sends the URL to the backend.
Backend checks if the website’s data already exists in the vector database.
---
5. If Data Exists
The app enables the question input field.
User can now enter a natural-language query.
The backend uses an AI model (e.g., GPT or Mistral) to generate and return an answer.
The answer is displayed in the chat interface.
---
6. If Data Does Not Exist
The backend starts scraping the website:
General text content
Contact information (emails, phone numbers, addresses)
Documents (PDFs, DOCs, etc.)
Scraping Progress UI
While scraping:
The app shows a progress screen with:
A loading spinner or progress bar
A message:
“Gathering information from the website. Please wait...”
Optional step updates:
“Scanning pages...”
“Extracting contact details...”
“Analyzing documents...”
“Finalizing...”
When scraping is complete:
A success message is shown:
“Data ready! You can now ask questions.”
The question input field is activated.
7. Asking a Question
User types a question.
Backend:
Searches the website's vectorized data.
Uses an AI model to answer.
The app displays the AI-generated human-readable response.
8. Data Maintenance
Re-scraping:
Each website is automatically re-scraped every 2 days to keep content fresh.
Auto-cleanup:
If a website’s data is not used for 30 days, it is deleted from the vector database.
9. History
The app maintains a list of all URLs entered by the user.
Accessible via the History tab in the top navigation bar.
Project Description
We are developing an Android chatbot application that answers user queries about any educational institute based on its website. The app should allow users to enter the URL of an institute’s website and then ask natural-language questions.
The backend will check whether the website’s data is already stored. If not, it will scrape the website for:
General text content
Contact information (emails, phone numbers, addresses)
Documents (PDFs, DOCs, etc.) with their contents extracted
The scraped data will be embedded and stored in a vector database. An AI language model (OpenAI, Mistral, or similar) will then use this data to generate human-like answers to user questions. If the data is already available, the user can immediately query it.
Additionally, the system must:
Maintain a list of URLs queried by each user
Automatically re-scrape and update the data for each URL every alternate day
Be crash-proof with robust error handling
The backend should be hosted on any free-tier platform (e.g., Render, Railway, Vercel).