Bindwell, an agricultural technology startup, has announced a $6 million seed funding round co-led by General Catalyst and A Capital. The funding will support the company's strategy to utilize artificial intelligence (AI) models for the in-house design of new pesticide molecules, with plans to license the resulting intellectual property directly to agrochemical firms. Y Combinator co-founder Paul Graham also contributed a personal investment to the round, which included participation from SV Angel.
Founded in 2024 by Tyler Rose, 18, and Navvye Anand, 19, Bindwell aims to adapt AI-led drug discovery techniques to agriculture to accelerate the identification and testing of novel pesticide molecules. The company's approach seeks to address current challenges in the agricultural sector, where pest resistance continues to grow, contributing to the loss of up to 40% of global crop production annually, according to the UN Food and Agriculture Organization. Traditional agrochemical methods often rely on modifying existing compounds, leading to increased chemical use without sufficiently combating evolving pests.
Bindwell initially entered Y Combinator's Winter 2025 batch with a business model focused on selling AI tools to established agrochemical companies. However, encountering resistance to adopting AI as a core component of pesticide discovery, the founders pivoted their strategy. Following a discussion with Paul Graham, they shifted to using their AI models to discover and develop new pesticide molecules internally. "The founders [of Bindwell] will probably do alright," Graham later posted on X. "They're smart and have a good idea."
The startup's AI suite includes Foldwell, a structure prediction model inspired by DeepMind's AlphaFold for identifying target protein structures; PLAPT, an open-source protein–ligand interaction model; and APPT, a protein–protein interaction model for biopesticide screening. Bindwell reports its APPT model outperforms existing tools by 1.7 times on the Affinity Benchmark v5.5. The system also integrates an uncertainty quantification mechanism to assess result reliability. Bindwell stated its models can analyze billions of molecules and deliver four times faster performance than DeepMind's AlphaFold 3.
Tyler Rose indicated the company focuses on identifying proteins unique to specific pests while being absent in humans or beneficial organisms. This target-based approach aims to design molecules that bind to and neutralize these pest-specific proteins. Bindwell is currently conducting efficacy tests of its AI-generated molecules at its San Carlos laboratory and is collaborating with a third-party partner for additional model validation. The startup is in early discussions with global agrochemical firms, anticipating its first licensing deal within the next year, and has initiated talks regarding field tests in India and China. Prior to this seed round, Bindwell secured pre-seed funding from Character Capital.
