Two teenage founders walked into Y Combinator co-founder Paul Graham’s backyard with an idea no one in agriculture seemed to want — an artificial intelligence model to help design better pesticides. By the time they left, they had a new business model, a new company, and ultimately, Graham’s backing.
Now, this redesigned company – Bidwell — raised $6 million in a seed round led by General Catalyst and A Capital, with a personal check from Graham himself. Instead of selling AI tools to legacy agrochemical giants, the startup is using its own models to design new pesticide molecules in-house and license IP directly — a shift in strategy aimed at modernizing a legacy industry still dominated by decades-old chemistry.
The use of pesticides in agriculture has doubled in the last three decadesstill up to 40% of the world’s crop production is still lost to pests and diseases every year, according to the UN’s Food and Agriculture Organization. As pests evolve and develop resistance, farmers are forced to use increasing amounts of chemicals just to maintain the same yields—a cycle that damages ecosystems and accelerates resistance even more. Regulatory pressure is increasing, but most agrochemical companies still rely on tweaks of legacy compounds. Bindwell is betting that AI can break the cycle by discovering entirely new, more targeted molecules—ones designed from the ground up for modern challenges.
Founded in 2024 by Tyler Rose, 18, and Navvye Anand, 19, Bindwell is adapting AI-driven drug discovery techniques to agriculture, aiming to speed up the way new pesticide molecules are identified and tested.
Bindwell began as a research project in late 2023, when Rose and Anand were students in the Wolfram Summer Research Program. They initially focused on an artificial intelligence model for drug discovery called PLAPT, which included binding affinity prediction—work that was later reported in a Nature Scientific Reports paper on cancer therapy. In 2024, they began investigating how the same approach could be applied to pesticides.
Both founders had personal exposure to the problem. Rose learned about the challenges of pest control from his aunt, who grows china. Anand, who has roots in Punjab, saw firsthand how limited pesticide options affected crop yields.
“Agriculture was on our minds,” Rose said in an interview. “That led to the realization that we can use exactly the same technology that was successful in drug discovery. We can transfer that to pesticide discovery because the biochemistry is the same, but pesticides are such a big problem and I feel like most people don’t focus on that.”
Techcrunch event
San Francisco
|
13-15 October 2026
Rose and Anand entered Y Combinator’s Winter 2025 batch with plans to build AI models and sell their access to large agrochemical companies. But they haven’t found traction—most industry players have been reluctant to adopt AI as a key part of pesticide discovery. Midway through the program, they were invited to Paul Graham’s home, where they talked with him for about 45 minutes on the back patio.
After hearing about their challenges, Graham suggested a different approach: Instead of selling tools, they could use their own models to discover new pesticide molecules themselves. This conversation marked the beginning of Bindwell’s current direction.
“The founders [of Bindwell] it will probably be fine,” he said later was posted at X. “They are clever and have a good idea.”
Bindwell has developed its own AI suite designed to reduce aliasing – a common problem where models produce unreliable or unsupported results. The software includes Foldwell, a structure prediction model, which is a custom diffusion system inspired by DeepMind’s AlphaFold, which is used to determine target protein structures. It also includes PLAPT, an open source protein-ligand interaction model capable of scanning every known synthetic compound in less than six hours, and APPT, a protein-protein interaction model for biopesticide screening, reported to outperform existing tools by 1.7× in Affinity Benchmark v5.5. Additionally, the suite incorporates an uncertainty quantification system that highlights when results are reliable and when more data is needed.
“Since we don’t sell AI models, we don’t compete with companies that sell models,” Rose told TechCrunch.
Together, Bindwell’s models can analyze “billions” of molecules, the startup said, and deliver performance four times faster than DeepMind’s AlphaFold 3.
“The way most pesticides are discovered right now is not target-based,” Rose said. “Entomologists and chemists propose different compounds and then test them on insects. You often need to synthesize and test thousands of chemicals, which is expensive just to test their effectiveness. With our AI models, you can simplify the problem to a single protein.”
Artificial intelligence helps identify proteins that are unique to a particular parasite but are absent in humans, beneficial insects or aquatic organisms such as water fleas.
“Once you find these proteins, you can design something that binds to them and prevents them from working,” Rose said.
Bindwell is currently testing the effectiveness of its AI-generated molecules in its San Carlos lab. It is also working with a third-party partner to further validate the models, though Rose declined to share details.
Rose said the startup is in early discussions with several global agrochemical companies, with its first partnership deal expected to close soon. “A year from now, we want to have licensing deals with some of these companies,” he said. Bindwell has also started talks with interested parties in India and China to conduct field trials.
The startup currently has a team of four and also works with external contractors to synthesize molecules.
Bindwell’s first round also included participation from SV Angel, along with Graham. Before joining Y Combinator’s Winter 2025 batch, the startup raised a pre-seed round from Character Capital.
