Command copied to clipboard
SYSTEMS & KNOWLEDGE ENGINE • STUDENT @ THAPAR

Local AI Struggle Tracker &
Lecture Study Engine

Notes-AI is a privacy-first desktop application built for computer science engineers. Log failing code attempts, auto-generate structured Markdown & PDF lecture notes, and query local study repositories with sub-millisecond SQLite FTS5 search.

Scanning environment...
• 100% Privacy & Offline Default • MIT Licensed
Download Desktop Package Read Architecture Spec →
in.sahuja.NotesAI — PySide6 Qt Desktop Shell
CORE MODULES
Struggle Tracker
Lecture Notes
Subject Master Guides
SYSTEMS & STORAGE
SQLite FTS5 Indexer
Google Drive Sync
OS Keyring Secrets
Search FTS5 index (query: "pointer boundary lock")... 0.42 ms
C++ / Systems Programming ID #104
Segmentation Fault in Linked List Base Case
Core Insight: Dereferencing NULL pointer before boundary check causes stack frame collapse. Fixed by ordering guard clause before member dereference.
Ollama Local Vision Model qwen2.5-vl:7b
Lecture Diagram Extraction Pipeline
Parsed 12 blackboard screenshots into OS_Lecture_04.md with embedded ASCII flowcharts and structured key takeaways.
ENGINEERING TRADEOFFS

Architecture & System Design

Engineered for strict data ownership, minimal runtime overhead, and cross-platform desktop reliability.

Struggle & Bug Tracker
Log failing code snippets, runtime error stack traces, and reference fixes. AI automatically categorizes issue types and synthesizes concise Core Insights.
SQLite FTS5 Full-Text Engine
Embedded zero-latency search indexing across struggles, markdown lecture notes, subject guides, and image metadata with instant query responses.
OS Keyring Secrets Store
Cloud API keys are stored securely inside operating system secret managers (GNOME Keyring, KDE Wallet, macOS Keychain, Windows Credential Manager) instead of plaintext files.
Multi-Provider AI Engine
Run 100% offline using local Ollama vision models (qwen2.5-vl / llama3), or connect OpenAI and Anthropic cloud backends.
Pure-Python PDF Pipeline
Custom fpdf2 compilation engine eliminating heavy GTK/WeasyPrint native system dependencies, ensuring seamless single-file binary distribution.
Google Drive PKCE Sync
OAuth 2.0 PKCE 1-click cloud synchronization to Google Drive with automated conflict resolution and local backup preservation.
DISTRIBUTION PACKAGES

Cross-Platform Installers

Native desktop packages for Linux distros and Windows. Free, open source, and lightweight.

Debian / Ubuntu (.deb)
Linux 64-bit • Native Deb Package
$ sudo dpkg -i notes-ai_1.0.0_amd64.deb
Download .deb Package
Fedora / RHEL (.rpm)
Linux 64-bit • Native RPM Package
$ sudo dnf install ./notes-ai-1.0.0.rpm
Download .rpm Package
Flatpak Package
Sandboxed • in.sahuja.NotesAI
$ flatpak install in.sahuja.NotesAI
View Flatpak Manifest
Windows Installer (.exe)
Windows 10 / 11 64-bit • Inno Setup
Notes-AI-Setup-1.0.0.exe
Download Windows .exe
PyPI / Pip Package
Cross-Platform • Python 3.10+
$ pip install notes-ai
View PyPI Source
DEVELOPER SETUP

Quick Start & Configuration

Get up and running in less than two minutes from terminal or source code repository.

1
Clone repository & create virtualenv:
git clone https://github.com/Sanyam-Ahuja/notes-ai.git && cd notes-ai && python3 -m venv .venv
2
Activate virtual environment & install editable package:
source .venv/bin/activate && pip install -e .
3
Launch desktop GUI:
notes-ai
1
Install & start Ollama service:
curl -fsSL https://ollama.com/install.sh | sh && ollama serve
2
Pull vision model for lecture screenshot parsing:
ollama pull qwen2.5-vl:7b
1
Build Debian package:
./build_deb.sh && sudo dpkg -i dist/notes-ai_1.0.0_amd64.deb
2
Build RPM package:
./build_rpm.sh