AI-Powered Pesticides: Teen Founders Raise $6M - TechCrunch

AI-Powered Pesticide Discovery: Bindwell Secures $6 Million Seed Funding

Two young entrepreneurs, while still teenagers, presented their innovative concept to Y Combinator co-founder Paul Graham. Their idea centered around utilizing an AI model to revolutionize the design of more effective pesticides.

This initial meeting led to a pivotal shift in their business strategy and ultimately, Graham’s investment. The company, now known as Bindwell, has successfully secured $6 million in a seed funding round.

A New Approach to a Legacy Industry

The seed round was co-led by General Catalyst and A Capital, with a direct investment from Paul Graham. Bindwell is diverging from the traditional model of selling AI tools to established agrochemical companies.

Instead, the startup is focusing on independently designing novel pesticide molecules using its proprietary AI models and directly licensing the intellectual property. This strategic move aims to modernize an industry heavily reliant on decades-old chemical compounds.

The Growing Need for Innovation in Pest Control

Over the past three decades, pesticide usage in agriculture has doubled. Despite this increase, up to 40% of global crop production remains vulnerable to pests and diseases annually, according to the UN Food and Agriculture Organization.

As pests develop resistance, farmers are compelled to use larger quantities of chemicals to maintain yields. This escalating cycle negatively impacts ecosystems and accelerates the development of further resistance. Increased regulatory scrutiny is emerging, yet many agrochemical companies continue to refine existing compounds.

Bindwell believes AI can disrupt this pattern by identifying entirely new, highly targeted molecules—designed specifically to address contemporary challenges.

Founders and Origins

Bindwell was founded in 2024 by Tyler Rose, 18, and Navvye Anand, 19. They are adapting AI-driven drug discovery techniques to the agricultural sector, aiming to accelerate the identification and testing of new pesticide molecules.

The company’s origins trace back to a research project initiated in late 2023, while Rose and Anand were students at the Wolfram Summer Research Program. Their initial work centered on PLAPT, an AI model for predicting binding affinity, which was subsequently featured in a Nature Scientific Reports publication concerning cancer therapeutics.

In 2024, they began exploring the application of this same methodology to pesticide development.

Personal Connections to the Problem

Both founders possess firsthand experience with the challenges of pest control. Rose gained insight from his aunt, a farmer in China, while Anand, with roots in Punjab, witnessed the impact of limited pesticide options on crop yields.

“Agriculture has consistently been a topic of interest for us,” Rose explained. “This led us to recognize the potential of applying the same technology that has proven successful in drug discovery to pesticide discovery. The underlying biochemistry is similar, but pesticides represent a significant problem that often receives less attention.”

From Initial Plans to a Pivotal Conversation

Rose and Anand entered Y Combinator’s Winter 2025 program intending to develop AI models and offer access to major agrochemical companies. However, they encountered limited interest, as many industry stakeholders were hesitant to integrate AI into their pesticide discovery processes.

A turning point came when they were invited to Paul Graham’s home for a 45-minute discussion on his patio.

Graham proposed an alternative strategy: instead of selling tools, they should leverage their own models to discover new pesticide molecules. This conversation fundamentally shaped Bindwell’s current direction.

Graham later commented on X, “The founders [of Bindwell] will probably do alright. They’re smart and have a good idea.”

Bindwell’s AI Suite

Bindwell has created a proprietary AI suite specifically designed to minimize hallucination—a common issue where models generate unreliable outputs. The suite includes:

• Foldwell: A structure prediction model, inspired by DeepMind’s AlphaFold, used to identify target protein structures.
• PLAPT: An open-source protein-ligand interaction model capable of scanning all known synthesized compounds in under six hours.
• APPT: A protein-protein interaction model for biopesticide screening, reportedly outperforming existing tools by 1.7× on the Affinity Benchmark v5.5.

The suite also incorporates an uncertainty quantification system to assess the reliability of results and identify when additional data is required.

“Because we are not selling AI models, we aren’t directly competing with companies that do,” Rose stated to TechCrunch.

Collectively, Bindwell’s models can analyze “billions” of molecules, achieving performance four times faster than DeepMind’s AlphaFold 3.

A Target-Based Approach

“Current pesticide discovery methods are often not target-based,” Rose explained. “Entomologists and chemists suggest compounds, then test them on insects. This often requires synthesizing and testing thousands of chemicals, which is costly, just to assess efficacy. Our AI models streamline this process by focusing on a single protein.”

The AI identifies proteins unique to a specific pest but absent in humans, beneficial insects, or aquatic organisms.

“Once these proteins are identified, a molecule can be designed to bind to them and disrupt their function,” Rose added.

Current Status and Future Plans

Bindwell is currently evaluating the effectiveness of its AI-generated molecules in its San Carlos laboratory. They are also collaborating with a third-party partner for further model validation, details of which Rose declined to disclose.

The startup is engaged in preliminary discussions with several global agrochemical firms, anticipating its first partnership agreement to be finalized soon. “Within a year, we aim to establish licensing agreements with some of these companies,” Rose said. Bindwell is also initiating conversations with stakeholders in India and China to conduct field trials.

The company currently comprises a team of four, supplemented by external contractors for molecule synthesis.

Investment Details

Bindwell’s seed round included participation from SV Angel, alongside Graham. Prior to joining Y Combinator’s Winter 2025 batch, the startup secured a pre-seed round from Character Capital.