Join waitlist
Back to products
PickleWatch

PickleWatch

PickleWatch on Apple Watch: Your ultimate pickleball partner

Website github.com
Overview

What it is

Shoutout Llamastack x 8VC challenge - WashU

Intent

I need it when

Prototype and test LLM-based desktop automation workflows locally before deployment

Provides an open-source, locally-hosted architecture (React frontend + Flask backend + LlamaStack) that developers can fork, modify, and extend to experiment with LLM-driven system control without vendor lock-in

Monitor and debug system command execution with visibility into what the LLM generated and how it performed

Displays command execution history with real-time status updates and includes a /status endpoint to check service and LlamaStack connectivity, helping users understand and troubleshoot LLM-generated commands

Automate repetitive macOS tasks through voice or text commands without writing scripts

Supports both text and voice input (via SpeechRecognition API) to trigger system commands, with real-time execution history and status feedback, enabling hands-free or quick-input task automation

Execute macOS system commands using natural language instead of memorizing terminal syntax

The product translates natural language input into Python code for macOS API calls via LlamaStack integration, allowing users to control their operating system by describing tasks in plain English rather than typing complex shell commands

Drop

Not a fit when

  • User requires a commercial, production-ready desktop application with vendor support and SLA guarantees
  • User needs cross-platform support beyond macOS (Windows, Linux native support required)
  • User requires enterprise security compliance, audit logging, and formal access controls
  • User needs a graphical user interface without command-line or code-level configuration
  • User operates in an environment where running local LLM inference servers is not feasible or permitted
Commercials

Pricing

Pricing not specified