Years ago, I walked into my first mega high-tech greenhouse and was stunned by how little they understood about what was happening in the facility. I asked what I thought were simple questions, what was the last activity performed in this row of fruit? How often were the plants tended? How much fruit would be produced by this row? How much by a single plant? How good was that worker, supervisor, or team? How many people does it really take to accomplish the work needed to run the facility? While people could in general explain the plan for what was needed, I was told it was “impossible” to monitor, track, or even attempt to answer the questions I was asking. I was told they would need an army of supervisors to track and audit all the rows. That they had no easy way to aggregate and visualize the data.
I would not let go and asked, “How do you know what you are doing is successful?”
The response… “My eye knows”. Not able to hold back my sarcasm I asked how I could remove their eye with a spoon and add it to a technology stack. (Drawing courtesy of my daughter who overheard my statement)
The more I dug in the more I realized that the tools, technology, and data just did not exist. That the “future” of agtech seemed to always focus on autonomous approaches that were capital intensive. They all painted a rosy picture that someday all farms would be fully autonomous. All this capital would pour in to support this idea, yet the prices of the commodity needed to stay low. Note that I was brought on board to design and build software to tell robotic harvesters what to do. I noticed that the viability of the robots I was working with were many years away, maybe even decades. There was a huge gap in how we go from where we are today to this wonderful future. I even had my doubts about this being a viable approach outside the most highly developed countries.
As I faced my new reality I started thinking of ideas and solutions that focused on empowering the labor of today over removing them. Trying to bridge the gap. On creating efficiency through visibility. Moving from the approach from a hyper focus on increasing yield to a focus on capturing the yield without waste. To track workers beyond how fast they worked, introducing quality as a critical metric. My first statement to Akash who became my cofounder was, “We must reduce plausible deniability, or we will never get efficiency. Create visibility where there is none. Surface where to act in the simplest ways.”
I equated what we needed to accomplish as intersection of quality and control navigating a curve in a NASCAR race. It is a delicate balance between speed and control. As you approach the turn at a high velocity you must maintain momentum without losing grip. Too fast risks the tires losing traction, sending you towards the outer wall, while going too slow sacrifices precious time. In my experience, all work being done whether it is white collar or blue has this intersection. Too fast and you introduce mistake. Too slow you introduce time. The key is finding and living on the edge of the intersection. The only way to do that is by being able to measure both quality and speed. Put a microphone in the forest, and you’ll know what a falling tree sounds like.
In my mind, we had to build a platform. One that focused on the optimal speed that ensured success. We needed to create visibility on which workers had mastered this intersection and which ones needed help. We needed to do this in a way that was globally useful and approach this through the AWS democratization of Software playbook. One that embraced all data and all approaches whether it was a worker with a phone, an IOT device, or even a robot all working together. We needed to build a language where anyone could describe their business, and we could run it without having to write more code.
The seeds of Tendrel were planted…