When Valeria Kogan received her Ph.D. program in bioinformatics, the scientific field that uses calculations and software to analyze biological data, in 2017, he envisioned his career would always be in mathematics, medicine, or biology. But after the first AI boom in the late 2010s, he found an interesting opportunity in a sector he hadn't considered: agriculture.
Kogan (pictured above, middle) told TechCrunch that when he was approached by a friend of a friend, a tomato farmer looking for someone who understood AI and could help apply the technology to agriculture, he was intrigued. When he heard that they were looking to use AI to monitor plant health, he realized that much of his bioinformatics background was transferable.
“When they started sharing their challenges with me, that plant health is there and their health is important, and how they monitor plant health, it was clear that technically the challenge is the same,” Kogan said. “You want to see and diagnose as quickly as possible.”
He decided to launch stop In 2020. The Tel Aviv-based startup uses computer vision and AI to monitor and diagnose greenhouse crops with diseases or pests. Fermata's software works with any off-the-shelf camera and takes pictures of the greenhouse crop twice a day. Its in-house AI model analyzes images and sends infection or disease alerts to farmers through an app.
Kogan admits that companies looking to bring AI to farms have struggled to gain meaningful traction in the past. While he doesn't think his company has gained meaningful market share yet, he thinks Pharmata's approach has allowed the group to gain traction for a few reasons.
For one, he said, they came into the market with genuine curiosity, to find out what greenhouse farms needed, as opposed to selling technology they didn't want.
“The original idea I had was to build a robot that would go through the greenhouse, and we built the first project,” Kogan said. “We made the first mistake, we built things before talking to anyone and it's still sitting in my dad's garage. When we started talking to people, it was very clear that nobody needs robots, that robots are a bad idea.”
He added that their method of training their AI models probably helped them stand out. Since the beginning, Pharmata has kept its data labeling team in-house as opposed to outsourcing, which Kogan attributes to the company's accuracy. When they started, they used publicly-available data, but now they train their models on their customers' data and have a research and development center where they test plants for various diseases.
“We love customers who have a lot of problems because it brings us a lot of data, especially if they have some serious disease,” Kogan says with a laugh. “We're like, OK, let's take the call seriously, and we're worried because it's bad news for them but it's amazing news for us.”
When Pharmata launched, Kogan said they thought it would make sense to partner with companies that were already selling to farms. That approach didn't garner much call back when Fermata started it in 2020, but that changed in 2022 when AI started gaining momentum. Now, Pharmata works directly with farms but also partners with large agricultural enterprises such as Bayer and Syngenta. The company declined to share growth metrics but has deployed more than 100 cameras.
Fermata recently raised a $10 million Series A round. The $10 million came entirely from Raw Ventures, a European VC. The firm was an existing investor, and Kogan said that when they set out to raise this current round, they didn't feel the need to dilute the cap table with more investors.
The round will be used to help the company scale and reach its goal of becoming profitable by 2026 Kogan says they've been able to grow this way without a sales team and largely from inbound interest, but their sales team is looking to grow. Farmata currently only deals in greenhouse grown tomatoes but is actively working towards expanding into new crops and increasing its shareholding.