The fall armyworm, a moth larva that can strip a cornfield bare, causes nearly $10 billion in crop damage in Africa alone [Perplexity Sonar Pro Brief]. It is also the first biological target for Bindwell, a San Francisco startup that is trying to apply the high-velocity, AI-driven logic of drug discovery to the slower, more stubborn world of pesticides. The company’s bet is not just on better algorithms, but on running the entire design-make-test loop itself, from digital screen to wet-lab validation, with the goal of licensing novel molecules to the agrochemical giants.
From software tools to full-stack discovery
Bindwell began with a more conventional idea: selling AI models to incumbent chemical companies. According to reporting, a conversation with investor Paul Graham helped pivot the strategy toward becoming a discovery engine in its own right [TechCrunch, November 2025]. The founders, Tyler Rose and Navvye Anand, decided they would not just predict molecules but also synthesize and test them, aiming to produce licensable intellectual property. This full-stack approach is the company’s central wedge. It asks agrochemical partners to buy a proven asset, not adopt a new, unproven software platform. The internal loop is powered by a trio of proprietary models the company has developed:
- Foldwell. A protein structure prediction model that the company claims runs roughly four times faster than DeepMind’s AlphaFold3, according to internal benchmarks [Bindwell, retrieved 2026].
- PLAPT. A compound-screening model built for speed. Bindwell states it can scan all 177 million known synthesized compounds in under six hours to find potential ligands for a target protein [The Teenagers Who Are Hacking Pesticides, retrieved 2026].
- APPT. An affinity prediction model for protein-protein interactions. The company reports it outperforms other models by 1.7x on the Affinity Benchmark V5.5 dataset, achieving a Pearson correlation coefficient (R) over 0.8 with experimental data [Introducing APPT: Affinity Protein-Protein Transformers, Bindwell, retrieved 2026].
The promise is to collapse a traditional pesticide discovery process that might require synthesizing and testing thousands of compounds down to just dozens [YC Teen Abandons Selling AI Tools to Agrochemical Giants, retrieved 2026].
The team that convinced General Catalyst
The founders’ youth is a conspicuous part of the narrative, but investors are betting on technical depth. Tyler Rose, the 18-year-old CEO, previously worked at Wolfram Research and has a background in machine learning and computational biology [Perplexity Sonar Pro Brief]. Navvye Anand, 19, left Caltech to work full-time as the scientific cofounder. They met at the Wolfram Summer Research Program in 2023 [Perplexity Sonar Pro Brief]. Their commitment was tested early: Wired reported that an investor offered them $750,000 to leave school and work on Bindwell full-time, and that they personally conducted early biological testing in Bay Area laboratories [Wired, October 2025]. That hands-on, proof-driven ethos appears to have carried into their seed round.
| Round | Date | Amount | Lead Investors | Key Participants |
|---|---|---|---|---|
| Seed | November 2025 | $6 million | General Catalyst, A Capital | SV Angel, Y Combinator, Paul Graham [TechCrunch, November 2025] |
The $6 million provides runway to expand a team that was reported as four people last November, and to fund the costly cycle of synthesis and biological validation [Villpress, November 2025]. The company is already hiring for lab technician and automation roles, signaling a commitment to building physical R&D capacity [Bindwell Careers, retrieved 2026].
The long road to a licensing deal
The ambition is clear, but the path is long and paved with technical and commercial risk. The most immediate question is whether the AI models’ impressive in-silico performance translates to real-world efficacy and safety in a crop field. Designing a molecule that strongly binds to a pest protein is one thing; ensuring it doesn’t harm beneficial insects, soil health, or human consumers is another, requiring years of regulatory testing. Furthermore, the business model hinges on convincing large, entrenched agrochemical companies to license an asset from a tiny startup. These firms have their own vast R&D divisions and are notoriously slow-moving. Bindwell has stated it is in early discussions with several global companies, but has not yet announced a signed partnership [TechCrunch, November 2025].
The market tailwind, however, is undeniable. Global pesticide use has doubled over the past three decades, yet about 40% of crops are still lost to pests and diseases, and resistance to existing chemicals keeps rising [YC Teen Abandons Selling AI Tools to Agrochemical Giants, retrieved 2026]. The demand for novel, targeted, and environmentally safer pesticides is a multi-billion-dollar problem waiting for a solution.
The unit economics of discovery
The math Bindwell must prove is not about software margins, but about the cost per viable candidate. If the traditional process costs $50 million and a decade to find one new pesticide, the startup’s proposition is to slash both figures by an order of magnitude. Their $6 million seed round is, in essence, a bet that their AI loop can find a promising molecule for a fraction of the historical cost. If they can get the cost of discovery down to, say, a few hundred thousand dollars per serious candidate, the licensing economics for big agchem become compelling. The incumbent they must beat isn’t another AI startup,it’s the internal R&D budget and timeline of a Syngenta or a Bayer. Bindwell’s real product isn’t a faster Foldwell; it’s a cheaper, faster path to a patent, and the clock is already ticking.
Sources
- [Perplexity Sonar Pro Brief] Bindwell company brief
- [TechCrunch, November 2025] Teen founders raise $6M to reinvent pesticides using AI, and convince Paul Graham to join in | https://techcrunch.com/2025/11/13/teen-founders-raise-6m-to-reinvent-pesticides-using-ai-and-convince-paul-graham-to-join-in/
- [Bindwell, retrieved 2026] Bindwell, Solving molecular discovery | https://bindwell.ai/
- [The Teenagers Who Are Hacking Pesticides, retrieved 2026] The Teenagers Who Are Hacking Pesticides
- [Introducing APPT: Affinity Protein-Protein Transformers, Bindwell, retrieved 2026] Introducing APPT: Affinity Protein-Protein Transformers, Bindwell
- [YC Teen Abandons Selling AI Tools to Agrochemical Giants, retrieved 2026] YC Teen Abandons Selling AI Tools to Agrochemical Giants, Turns to Pesticide Sector and Secures $6 Million in Funding
- [Wired, October 2025] How to Build an AI Startup: Go Big, Be Strange, Embrace Probable Doom | https://www.wired.com/story/ai-startup-interview-founders/
- [Villpress, November 2025] Bindwell: Teen Founders Raise $6M to Reinvent Pesticides with AI, and Win Paul Graham’s Support | https://villpress.com/bindwell-teen-founders-raise-6m-to-reinvent-pesticides-with-ai-and-win-paul-grahams-support/
- [Bindwell Careers, retrieved 2026] 实验室技术实习生, 百恩德 Bindwell | https://bindwell.ai/zh-cn/careers/lab-technician-intern