DEV Community: Ankush Chakraborty

NoteWise_AI

Ankush Chakraborty — Fri, 16 Oct 2026 13:10:48 +0000

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

What I Built

I built NoteWise AI, a local-first AI study assistant designed to help students study more efficiently using their own notes.

Students often have to go through lengthy PDFs and lecture notes to prepare for exams. NoteWise AI aims to make this process easier by allowing students to upload their notes and use AI to:

Generate short summaries and exam-oriented revision notes.

Ask questions about their study material.

Generate practice quizzes and answer keys.

Download generated summaries and quizzes for revision.

The initial version supports text-based PDF and TXT files.

The project is intended for students, independent learners, and tutors. In the future, it could become a low-cost study tool for students who want useful AI features without paying for an expensive subscription.>

Code

ANILION-CODE / NoteWise_AI

A local first study assistant that uses an open*weight language model through Ollama to summarize notes, answer questions from uploaded material, and generate practice quizzes.

NoteWise AI

A local first study assistant that uses an open*weight language model through Ollama to summarize notes, answer questions from uploaded material, and generate practice quizzes.

Features

Upload selectable*text PDF or TXT notes
Generate short, detailed, or exam revision summaries
Ask questions using a lightweight keyword*based context selector
Generate practice quizzes and download results as Markdown
Run inference through Ollama on your own computer

Requirements

Windows 10/11
Python 3.10 or newer
Ollama for Windows
Enough RAM to run the model you choose

Run locally

Install Python from https://www.python.org/downloads/ if you do not already have it.
Install Ollama from https://ollama.com/download.
Open Command Prompt or PowerShell and download the default model:
```
ollama run llama3.2:3b
```
Wait for the model to download. When the model prompt appears, you can type /bye to exit. Leave Ollama running in the background.
Extract this repository, open a terminal in the notewise_ai folder, and…

View on GitHub

How I Built It

I built NoteWise AI using Python and Streamlit, with Ollama providing access to a locally running open-weight language model.

The main technologies are:

Python: Application logic and document processing.

Streamlit: Browser-based user interface.

Ollama: Runs the language model locally.

Llama 3.2 3B: The default model configuration for generating summaries, answers, and quizzes.

pypdf: Extracts text from PDF documents.

Lightweight keyword retrieval: Selects relevant text from notes when answering questions.

The application sends prompts to the local Ollama service instead of relying on a hosted AI API. The initial version uses a simple retrieval approach, which I plan to improve with embeddings, vector search, and source references.

Why Does Open Innovation Matter?

Open innovation is important to NoteWise AI because students should have more choices in how they use AI tools and where their data is processed.

Using an open-weight model through Ollama makes local inference possible. In the intended local setup, study documents and prompts can remain on the user's computer rather than being sent to a third-party AI API.

It also gives the project flexibility. Developers can inspect and modify the code, experiment with compatible models, and improve retrieval or prompting for different subjects.

Running a model locally can reduce recurring per-request API charges, although hardware, electricity, and setup still have costs. This could make NoteWise AI useful for students who want to avoid depending on paid AI subscriptions.

The goal is to build a tool that users can understand, adapt, and improve instead of locking the entire application to one proprietary AI provider.

GreenBuddy

Ankush Chakraborty — Sun, 11 Oct 2026 09:53:40 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

GreenBuddy is an AI-powered gardening assistant designed to help beginners turn gardening questions into real-world action.

Users can enter their city, available growing space, sunlight conditions, and the plants they want to grow. GreenBuddy generates a personalized gardening plan with practical care tips and an outdoor mission to complete.

The idea is simple: spend less time looking at a screen and more time growing something in the real world

Code

ANILION-CODE / GreenBuddy

GreenBuddy is an AI-powered gardening assistant designed to help beginners turn gardening questions into real-world action. Users can enter their city, available growing space, sunlight conditions, and the plants they want to grow. GreenBuddy generates a personalized gardening plan with practical care tips and an outdoor mission to complete.

How I Built It

GreenBuddy is built with Python and Streamlit, with Hugging Face Inference powering its AI gardening assistant.

The prototype collects basic information about a user's growing environment and sends it to an open-weight language model to generate personalized gardening guidance.

The application is designed to be beginner-friendly, with a simple interface and practical recommendations instead of an unnecessarily complicated setup.

Why Does Open Innovation Matter?

Open innovation makes it possible to experiment with AI without depending entirely on proprietary models or building expensive infrastructure.

By using an open-weight model through hosted inference, GreenBuddy can explore model choice and customization while keeping the application code accessible to other developers.

The project also leaves room for future improvements, including alternative models, local inference, and a gardening journal that users control.

My goal is to make AI useful beyond the screen: helping people learn about plants, spend time outdoors, and build a gardening habit.

first post

Ankush Chakraborty — Thu, 01 Oct 2026 12:21:52 +0000

this must be done