4.5.23-AI-Aviant

IAN SEIFERLING: I'm just going to shorten my pitch and say we do what [? Leela ?] does, but for controlled farms. That's it. But essentially, this is a really great talk. Really exciting to see that. And essentially, we're taking AI applications to turn controlled farms into hyperefficient lean factories using plant-level data.

I'm the CEO. My name is Ian Seiferling. Building AdaViv with a great team. Julian Ortiz, one of my cofounders, ex McKinsey, MIT MBA. Our CTO and cofounder Moe Vazifeh, quantum physicist, AI engineer. We work together as research fellows at lab at MIT. So we spun out here. We were incubated here. Now based out of Greentown Labs in Somerville.

So farming-- really, really tough job. We all know the challenges. Crop losses-- massive, massive losses today. We're losing 30%, 40% of crops annually. At the same time, these farmers are really challenged with high costs of labor, high costs of inputs. So it's really, really hard to drive high margins in these crops. And this is obviously a really important industry for society.

Controlled-environment farms-- no different, really. They do allow us to protect against the weather, to be more precise with the inputs. But those growers still struggle with visibility on the farm-- running around putting out fires. They spend much of their time on administrative tasks. And to compensate-- or to overcompensate-- that is they often end up overspending on inputs and labor.

So what we're doing at AdaViv is building a lean cultivation platform that puts eyes on all their plants and uses that data to help them optimize those workflows, to detect and react rapidly, to be more targeted, and to optimize their productivity so that every grower and technician is handling more plants per person, ultimately driving yields higher and reducing input costs.

How we do it-- we've built, really, an end-to-end platform that leverages machine vision, AI, smart in-field apps, and integrations with the farm data and the farm systems. So we image each plant with our machine vision technology. They move around the farms imaging different visual and nonvisual spectra. With that data, we essentially index the farm. And then we build models to detect growth, to detect stress, pests, disease, and tie that to all the SOPs and all the labor and activities that are going on.

So at the end, we're helping them create digital SOPs to target treatments faster, to optimize the level of treatments, and to make better cropping decisions based on all those inputs that they do have to control for, which I'll speak to.

We, today, have kind of focused on the areas that are causing the most pain for these farmers and these growers. Plant health is really the big one that we first focused on. We've helped them mitigate losses from pests and diseases like botrytis, viruses, powdery mildew. These are big ones that affect many crops. Again, we use that plant-level data and workflow software to help them boost labor productivity. So they're doing a lot to these plants-- touching them, deleafing, treating. And through these digital tools that we've built, we can essentially increase the number of plants that every technician is able to work on.

Irrigation-- big, big important inputs and big, big problems. Those systems fail all the time. And it's really hard to optimize, at a cultivar level, how to create that optimal irrigation strategy. So we've used thermal data that we collect to help them understand how the plant is reacting, detect failures in those systems early so that they can mitigate those losses. And much the same with lighting-- another, obviously, one of the big inputs into crop growth. So again, we help identify equipment failure, which happens surprisingly a lot, and also optimize those recipes.

What we're all talking about here today-- partnerships, expansion, ecosystem building-- these are all very important in agriculture and equally important for us in our business strategy. We are looking to expand crops, to expand geographies, leafy greens in the US, flowers in Latin America. And I think, as this type of technology gets more embedded in farms, we really have to build an ecosystem. It's the only way to be able to tie all these fragmented pieces of technology together. So we're very excited about that because we can use our plant-level data to really optimize, and automate, and make those inputs, those controls, the lighting, the crop protection strategies, much more informed, much more adaptable, much smarter.

So the type of companies we want to do those type of integrations with-- agroscience, crop protection, suppliers in the agriculture industry-- but also those suppliers of those fundamental core technologies-- lighting, HVAC system, control systems-- and also folks that can help us improve the impact and scalability of our own products-- so imaging, sensing, manufacturing, and automation. So super excited about that. If you're interested in any of those areas, again, please come track me down in the room. I'll be happy to talk.

And it's an incredibly important challenge. And like I said, the only way we're going to solve it is through building an integrated ecosystem. So thank you and talk soon.