Voice to Database

Your farm records can be kept simply, via voice message

Last season’s tomatoes in Field 3 were stunning — your best crop ever. This year? They’re struggling, and you can’t quite remember what you did differently. Was it the fish emulsion timing? The variety? The cover crop you planted? The answers are probably scattered between your phone’s notes app, that stack of receipts in your office, and the mental notes you told yourself you wouldn’t forget.

Every day on your farm, you’re gathering crucial information — soil test reports, sap analysis, pest patterns, variety performance, weather impacts. But these pieces of data end up scattered everywhere: pdf reports here, receipts there, mental notes that slip away, and voice memos that never get listened to after recording. Maybe you have an amazing memory and can piece it all together like Sherlock Holmes. Or maybe you’re like me, and you tell yourself that of course you’ll remember spraying that extra fish emulsion on the first three rows in field three, only to wonder a few weeks later why those peppers are doing so well.

At Good Agriculture, we believe there has to be a better way. For the past six months, we’ve been exploring a simple idea: what if you could just speak your observations as you work, and have them automatically organized into usable data?

Consider this: “Aphids on the romaine in high tunnel four.”

Two months earlier: “Poor germination on the romaine, power went out, and germination chamber appears to have overheated. Seeded four additional trays and planted out what did germinate in high tunnel four, row one.”

Could these two be connected? Looking at it on paper, it seems like there might be a connection, but would you remember that if you didn’t have an easy way to capture and compare that data?

Working with nine farmers near Atlanta, we’ve been testing this concept. The premise is simple: farmers send us voice memos about their daily activities. We transform these into structured data that can reveal patterns and connections that might otherwise go unnoticed.

Even in our early testing, we’re seeing hints of how powerful this can be. Check out Table 1 — this example is from one of our farmers in North Georgia.

Table 1. Bed 2, Field 1

This seems like a basic log until you consider that everything in that table came out of voice memos — the farmer didn’t have to write anything down, enter it into any programs or try to remember what he did on a given day. And this is searchable — next year, this farmer will be able to refer back to what he grew in this bed at any point throughout the year and make decisions on crop rotations, amendments or cover crops.

Table 2 demonstrates another example from a different farmer in our trial.

Table 2. Caterpillar Tunnel #1

Even from just these early trials, patterns are emerging that would be easy to miss in traditional record-keeping. That Spectrum application? Next spring, the farmer can compare soil test results to see if it actually boosted microbial activity. The succession planting timing in the caterpillar tunnel? It’s now documented for fine-tuning next season’s schedule. Each voice memo builds a more complete picture of the farm’s story, and we can match it with additional records like weather and day length to expand the potential insights.

We see enormous promise in this approach, but we need your help to build something truly useful. Imagine:

• Planning next season’s crops with field-by-field history at your fingertips: soil test trends, amendment timing, and crop performance all in one place
• Simplifying organic certification with automated record compilation from your voice notes and reminders to capture missing data
• Getting timely reminders based on your own patterns: “It’s been 20 days since those tomato transplants went in — time to check growth, cultivate and side-dress”
• Discovering patterns across seasons: which varieties consistently perform best in which fields, how different cover crops affect subsequent crops, or why certain pest problems keep recurring in specific areas

This isn’t about replacing the farmer’s knowledge — it’s about helping you capture and use more of what you already know. We’re looking for input from all farmers, but we’re especially eager to work with annual vegetable growers who might want to help shape the next phase of development and test our early prototypes.

We want to understand:

• What information do you wish you could capture more easily?
• What connections would be most valuable to your operation?
• How would you use this kind of tool in your daily work?

Ready to help build something better? Share your thoughts at goodagriculture.com/DFH-ideas. We’d love to follow up with a conversation about what would be most valuable for your operation.

Kirsten Simmons is the co-founder and Chief Farming Officer for Good Agriculture (goodagriculture.com), a company dedicated to helping farmers find funding, manage finances, reach new customers and get certified. She also farms at Ecosystem Farm in Atlanta (ecosystemfarm.com).