Local AI-powered office automation for people who work with files, documents and data every day.
No cloud. No subscription. Runs entirely on your machine via Ollama.
Office workers spend hours on repetitive tasks: renaming files, extracting data from documents, summarising reports, turning meeting notes into action items. Office AI Studio automates all of this - with or without AI - from a single desktop app built on Python and Ollama.
Everything runs locally. Your files never leave your machine.
⬇ Smart Drop Zone - drag any file onto the app and instantly choose what to do: summarize, extract data, translate, rename, copy, hash, and more. Works on multiple files at once.
⚡ Pipeline Builder - build multi-step AI workflows visually. Each step passes its output to the next. Drag files directly onto step cards. Save and reload named pipelines.
» Scripts - write, save and run Python automation scripts. Five starter scripts included. AI can generate a new script from a description or explain any existing one.
⏲ Auto Tasks - create automation rules based on file patterns. Drop files onto a task card to run it instantly. Tracks run count and last-run time.
⊞ Data Tools - open, preview, merge, clean and export CSV files without Excel. AI can analyse the data and describe patterns or quality issues.
◎ Meeting Notes - paste raw notes and get three outputs: action items with owners, executive summary, and a ready-to-send follow-up email draft.
≡ File Manager - browse the filesystem with a text preview pane and right-click context menu.
◉ Chat - multi-turn conversation with any local Ollama model.
📝 Notepad AI - full-screen editor with an AI assistant panel. Insert or replace text with AI output.
>_ Terminal AI - real shell with ?? <question> to ask AI and an auto-explain mode for command output.
# 1. Install Ollama
# https://ollama.com/download
# 2. Pull a model
ollama pull llama3.2:3b
# 3. Start Ollama server
ollama serve
# 4. Install Python dependencies
pip install -r requirements.txt
# 5. Run
python office_ai_studio.py| Package | Required | Purpose |
|---|---|---|
requests |
✅ Yes | Ollama API communication |
tkinterdnd2 |
Recommended | Drag & drop file support |
chardet |
Recommended | Auto-detect file encoding |
python-docx |
Optional | Read .docx files |
openpyxl |
Optional | Read .xlsx files |
pandas |
Optional | Advanced data analysis |
tkinter is bundled with Python on Windows. On Linux:
sudo apt install python3-tkThe app tries /api/chat first and automatically falls back to /api/generate for base models without a chat template.
| Model | Status |
|---|---|
llama3.2:3b |
✅ |
llama3.2:1b |
✅ |
SpeakLeash/bielik-*-instruct |
✅ |
mistral:* |
✅ |
qwen2.5:* |
✅ |
| Base models (no chat template) | ✅ generate fallback |
All data is stored locally in ~/.office_ai_studio/:
| File | Contents |
|---|---|
pipelines.json |
Saved pipelines |
history.json |
AI run history (last 500 entries) |
scripts.json |
Saved automation scripts |
tasks.json |
Auto task rules |
| Shortcut | Context | Action |
|---|---|---|
Ctrl+Enter |
Chat, Notepad AI | Send message |
F5 |
Scripts | Run script |
↑ / ↓ |
Terminal AI | Navigate command history |
?? prefix |
Terminal AI | Ask AI a question |
| Double-click | File Manager | Open file / enter folder |
Tomasz Wietrzykowski
GitHub · X / Twitter · WebSim
MIT - free to use, modify and distribute.
