Document Understanding Assistant

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

📄 AI Document Understanding Assistant

I built an AI-powered Document Understanding Assistant using open-source AI technology. It helps users understand text documents by generating summaries and answering questions about their content.

✨ Key Features

• 📄 Load and read text documents.
• 📝 Generate summaries automatically.
• 💬 Ask questions about a document and get AI-generated answers.
• 🖥️ Use a locally running AI model through Ollama.

🛠️ Who is it for?

This project is useful for students, researchers, and anyone who needs to understand documents quickly without reading every line manually.

💡 The goal: Make document understanding easier and faster using local AI, without depending on a paid cloud AI API.

Code

Varshita12345 / document-assistant

AI Document Assistant

A Streamlit app for summarizing text documents and asking questions about them with the Gemini API. The command-line main.py remains available for local Ollama use.

Run locally

1. Install the dependencies with pip install -r requirements.txt.

2. Create .streamlit/secrets.toml with:

   GEMINI_API_KEY = "your-Gemini-API-key"

3. Start the web app with streamlit run app.py.

Keep the API key private. .streamlit/secrets.toml is ignored by Git.

Deploy on Streamlit Community Cloud

1. Push this repository to GitHub.

2. Sign in at Streamlit Community Cloud with GitHub and choose Create app.

3. Select Varshita12345/document-assistant, the main branch, and app.py as the app file.

4. Before deploying, open Advanced settings / Secrets and add:

   GEMINI_API_KEY = "your-Gemini-API-key"

5. Deploy. Streamlit will display the public app URL when deployment succeeds.

View on GitHub

How I Built It 🛠️

I built my AI Document Understanding Assistant using Python and open-source AI technologies.

🔧 Technologies Used

• Python — for the application logic.
• Ollama — to run the AI model locally.
• Gemma (gemma4:e2b) — for document summarization and question answering.
• Streamlit — to build a simple web-based user interface.
• VS Code — for writing and testing the code.
• Git and GitHub — for version control and sharing the project.

⚙️ How It Works

1. The user provides a text document.
2. The application loads the document content.
3. The content is sent to the locally running Gemma model through Ollama.
4. The AI generates a summary or answers questions about the document.
5. The result is displayed to the user.

I first built and tested the document-loading, summarization, and question-answering features in Python. I then started working on a Streamlit interface to make the application easier to use.

🌱 What I Learned

This project helped me explore local AI inference, open-source AI models, Python development, and building practical AI-powered applications.

My goal is to keep improving the project and make document understanding simpler and more accessible.

Why Does Open Innovation Matter? 🌍

Open innovation matters because it makes AI development more accessible to students, independent developers, and small teams.

For my AI Document Understanding Assistant, using Ollama and the Gemma model allowed me to experiment with local AI inference without relying on a paid, hosted AI API.

It gave me the opportunity to:

• 💻 Explore how AI models work locally.
• 🔓 Build with open AI tools and learn from their ecosystems.
• 💰 Experiment without paying per request to a cloud AI API.
• 🔐 Keep document processing local when the application is configured to use only the local model.
• 🚀 Build and improve a practical AI project using accessible technologies.

As a student developer, open innovation helps turn ideas into working prototypes while building real-world technical skills.

I believe open innovation makes AI more accessible and empowers more people to create, experiment, and contribute to technology.

My Agent Session

My First User 👩‍💻

My first user was me! 🎉

I tested my AI Document Understanding Assistant using a sample text document. I tried the document summarization and question-answering features to check how well the local Gemma model could understand the content.

Seeing the application generate a summary and answer questions was an exciting milestone. As a student developer, this hands-on testing helped me understand how open-source AI models can be used to build practical applications.

My next goal is to let other students try the assistant, collect their feedback, and improve the user experience. 🚀

Prize Categories

Prize Categories 🏆

• Best Use of Gemma — Using the Gemma open-weight AI model through Ollama for local document summarization and question answering.>

i am the project submitter and my team member is divyagunda97-crypto