This post covers three weekend projects addressing specific technical problems: task management with resource limits, AI assistant integration, and market microstructure analysis. Each project uses FastAPI and Python, prioritizing straightforward implementations over comprehensive feature sets.

1. NanoSaaS: Task control with resource quotas

A lightweight system for managing computational tasks with credit-based resource limits.

Features:

  • Credit system for resource allocation and abuse prevention
  • Real-time task monitoring without page refreshes
  • Google SSO authentication
  • Stack: FastAPI, Celery, Alpine.js, Tailwind CSS

The design targets small teams or individual developers who need task tracking with quota enforcement.

GitHub: https://github.com/carlosplanchon/nanosaas/

2. FastAPI AI Assistant: OpenAI integration template

A minimal FastAPI application that wraps OpenAI’s Assistant API with streaming responses.

Features:

  • Streaming responses via Server-Sent Events (SSE)
  • FastAPI middleware layer between OpenAI and frontend
  • Stack: FastAPI, Alpine.js, Tailwind CSS

The application serves as a template for integrating conversational AI into existing systems.

GitHub: https://github.com/carlosplanchon/fastapi_ai_assistant

3. MeasureVolume: Order book analysis

Analyzes order book snapshots to estimate trading volume and liquidity changes in cryptocurrency markets.

Features:

  • Order book diffing to infer executed trades
  • Sample datasets for testing
  • Market taker activity measurement scripts

The tool provides a lightweight approach to studying market microstructure.

GitHub: https://github.com/carlosplanchon/measurevolume