Garden Party: Using LLMs for OCR and Data Analytics
Unlocking the Secrets of Garden Productivity
In this article, I’ll share my experience using large language models (LLMs) to digitize handwritten garden yield data and perform analytics, exploring the intersection of technology and gardening.
Introduction
Several years ago, I wanted to start a small garden. The problem? I didn’t know anything about gardening and didn’t have much space. After doing a bit of research, I started using the Square Foot Gardening system, invented in 1975 by Mel Bartholomew. Being a former engineer and efficiency expert, Mel wanted to solve the problems of inefficiency and wastefulness in single-row gardening. Essentially, the square-foot garden method is a high-density gardening method that uses about 20% less space than a traditional row garden and 10% of the water.
Square foot gardening involves creating 4'x4' raised beds divided into sixteen 1'x1' squares, with specific planting formulas for each square. For example, within each 1’x1’ square, you can plant:
● 16 radishes
● 9 onions
● 4 lettuce plants
● 1 tomato plant
