Notion Workflows

Automated workflows for managing and enriching Notion databases with AI-powered content generation.

Overview

A collection of automated workflows that manage and enrich Notion databases — from meal planning and shopping lists to travel guides and pub crawl routes. The system runs nightly via GitHub Actions, using TypeScript, the Notion API, and Google Gemini AI to automate repetitive tasks and generate intelligent content.

Key Workflows

Manage Shopping

End-to-end shopping list manager that extracts ingredients from meal plans and uses Gemini AI to categorize items automatically.

Populate Films & TV

Enriches a films/TV database with metadata from TMDB and OMDb APIs — posters, overview, runtime, genres, directors, and ratings.

Populate Meals

Uses Gemini AI to auto-complete meal recipes with ingredients and cooking instructions following British English conventions.

Populate Pubs

Calculates optimal pub crawl routes using geolocation, Haversine formula, and nearest-neighbour optimization.

Manage Meal Planner

Maintains a rolling 28-day meal planner by archiving past days and creating future day pages automatically.

Events Calendar

Generates recurring calendar events from templates, creating events for the next 12 months based on weekday and week ordinal rules.

Technologies

TypeScriptNode.jsNotion APIGoogle Gemini AIGitHub ActionsTMDB APIOMDb API

Project Structure

utils/
  ai.ts          # Gemini AI client and batch annotation helper
  logger.ts      # Structured logger with log levels
  notion.ts      # Notion client, page/block CRUD helpers
  parsing.ts     # Property builders for Notion API requests

workflows/
  cleanup-shopping/          # Remove checked items from shopping lists
  events-calendar/           # Generate recurring calendar events
  manage-meal-planner/       # Rolling 28-day meal planner
  manage-shopping/           # AI-powered shopping list management
  populate-films-and-tv/     # TMDB/OMDb metadata enrichment
  populate-meals/            # AI-generated recipes
  populate-pubs/             # Optimal pub crawl routing
  populate-travel-database/  # AI travel destination annotations
  populate-walks-database/   # AI walking location annotations

Technical Highlights

Automated CI/CD: All workflows run on nightly cron schedules via GitHub Actions with manual trigger support
AI Integration: Leverages Gemini 2.5 Flash with JSON response mode for structured content generation
Geospatial Algorithms: Implements Haversine formula and nearest-neighbour optimization for route planning
API Orchestration: Integrates multiple external APIs (Notion, TMDB, OMDb, Gemini) with robust error handling
Modular Architecture: Shared utilities for logging, parsing, and API clients enable rapid workflow development

← Back to Projects

View on GitHub →