SnapCapture

SnapCapture is a lightweight operational capture tool for quickly collecting tasks, reminders, appointments, and follow-ups from text, screenshots, PDFs, images, and audio recordings.

The application is designed as a temporary capture buffer — not a permanent organizer.

Users can:

• capture information quickly,
• review extracted items,
• edit titles, dates, and notes,
• export tasks and events,
• remove items afterwards.

---

Features

• Text extraction
• PDF and image extraction
• Audio transcription
• AI-assisted task and event extraction
• Multiple extraction engines
• Temporary operational queue
• .ics calendar export
• Mobile-friendly interface
• Daily AI request limits
• Privacy-focused workflow
• BYOK (Bring Your Own Key) support

---

Current MVP Status

SnapCapture is currently an MVP release focused on:

• operational workflows,
• fast capture,
• simple review,
• temporary storage,
• export-first usage.

The application is intentionally lightweight and does not try to replace calendars, organizers, or note-taking systems.

---

Workflow

Capture → Extract → Review → Save → Export → Delete

Recommended usage:

• collect tasks and events during the day,
• review extracted items,
• export them to your preferred calendar or organizer,
• delete them afterwards.

---

Supported Inputs

• Text
• Screenshots
• Images
• PDFs
• Audio recordings

---

Extraction Engines

SnapCapture currently supports:

• Gemini Flash
• GPT-4o-mini

Additional providers may be added later.

---

BYOK (Bring Your Own Key)

Advanced users can optionally configure their own API keys.

This allows:

• higher usage limits,
• provider flexibility,
• personal API billing,
• reduced shared infrastructure costs.

SnapCapture does not permanently store user API keys.

---

Privacy

SnapCapture is designed with a privacy-focused workflow:

• local OCR and transcription whenever possible,
• temporary storage only,
• export-first philosophy,
• no long-term operational archive.

Uploaded files should not be stored permanently.

---

AI Usage

AI extraction is used only when needed.

Typical flow:

1. Local OCR/transcription
2. AI cleanup and extraction
3. Structured operational output

This helps reduce API usage while improving extraction quality.

---

Daily AI Limits

The MVP currently uses daily AI request limits to prevent excessive API usage.

Current default:

• 30 AI requests per day

---

Technology Stack

Frontend

• React
• Vite

Backend

• FastAPI
• Python

Database

• SQLite

Local Processing

• Tesseract OCR
• Whisper transcription

AI Providers

• Gemini
• OpenAI

---

Installation

Prerequisites

Install:

• Node.js
• Python 3.11+
• Tesseract OCR

Optional:

• FFmpeg (recommended for audio support)

---

Frontend Setup

cd frontend

npm install

npm run dev

Frontend runs on:

http://localhost:5173

---

Backend Setup

Create virtual environment

cd backend

python -m venv .venv

Activate environment

macOS / Linux

source .venv/bin/activate

Windows

.venv\Scripts\activate

---

Install dependencies

pip install -r requirements.txt

---

Run backend

uvicorn app.main:app --reload

Backend runs on:

http://localhost:8000

---

Environment Variables

Create .env inside backend directory:

GEMINI_API_KEY=your_key
OPENAI_API_KEY=your_key

Never commit .env files to GitHub.

---

OCR Setup

Install Tesseract OCR.

macOS

brew install tesseract

Ubuntu

sudo apt install tesseract-ocr

Windows

Download installer from: https://github.com/tesseract-ocr/tesseract

---

Audio Support

Optional but recommended:

macOS

brew install ffmpeg

Ubuntu

sudo apt install ffmpeg

---

Notes

SnapCapture is still under active development.

Current focus areas:

• extraction quality,
• mobile workflow,
• AI provider flexibility,
• .ics and .csv export,
• operational simplicity.

---

Version

SnapCapture MVP v0.9

BKlein Digital Labs

Small practical tool for real everyday usage.

---

© 2026 BKlein Digital Labs. All rights reserved.# SnapCapture