Company profile · 4 min read
Bindwell: An in-house design–make–test loop for safer pesticides
Bindwell uses AI models and wet-lab feedback to discover pesticide molecules, beginning with target-specific agrochemical research.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Bindwell is building new pesticides in-house with AI models and wet-lab feedback. It fits agrochemical partners and investors interested in molecule discovery where target specificity, resistance and ecological safety matter more than generating another plausible compound.
What it does
Bindwell’s current site describes a system that designs a molecule, makes it and tests whether it works in a living organism. The company starts with pesticides because the feedback loop is shorter than in many pharmaceutical programs. Its public model stack includes Foldwell for structure prediction, PLAPT for protein-ligand interaction and APPT for protein-protein interaction and biopesticide screening.
The YC launch says Foldwell runs four times faster than AlphaFold 3, PLAPT can scan known synthesized compounds in under six hours and APPT outperforms an existing model by 1.7x on Affinity Benchmark V5.5. Those are company-reported model claims; they are not independent benchmarks or proof that a candidate works in the field. The more interesting product choice is that Bindwell is building its own compounds and intends to license the resulting IP, rather than selling a general-purpose model to agrochemical companies.
The current site says the team can move from design to a real organism-level answer in a day. That is a company claim about its internal loop, not a guarantee across species, formulations or regulatory stages. The buyer should care about which pest, target and assay the system has already validated and how uncertainty changes the next experiment.
Founder context and tradeoffs
The YC profile identifies Navvye Anand and Tyler Rose as founders. It describes Navvye’s bioinformatics and AI interest and Tyler’s machine-learning and computational-biology background at Wolfram Research. TechCrunch reported that the founders shifted from selling models to designing pesticide molecules themselves after early agrochemical companies were reluctant to adopt AI as a core tool, and reported a $6M seed round. That article is context, not an independent validation of the science.
Pricing is not public. The visible model is partnership and IP-led rather than self-serve software. A potential partner should ask about ownership, field efficacy, non-target effects, resistance, synthesis and regulatory evidence before treating model speed as commercial value.
Editorial take
I would shortlist Bindwell for a focused agrochemical partnership where a specific pest and biological target can produce fast experimental feedback. I would not evaluate it by a generic AI demo. The proof is a repeatable design–make–test loop that produces safer, more selective candidates.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | In-house AI and wet-lab system for pesticide discovery |
| Buyers | Agrochemical partners, researchers and eventual IP licensees |
| Public model claims | Foldwell, PLAPT and APPT speed/benchmark claims, company-reported |
| Pricing | Not publicly listed; partnership-led |
| Main gate | Living-organism validation, selectivity, resistance and regulatory evidence |
Sources checked
| Source | Checked |
|---|---|
| YC profile | 2026-09-19 |
| Bindwell homepage | 2026-09-19 |
| Bindwell launch | 2026-09-19 |
| TechCrunch coverage | 2026-09-19 |
Cohort context
Bindwell is listed in Winter 2025. In our 2026-09-18 directory snapshot, 19 of 165 listed companies in that cohort have YC’s primary industry label Industrials (11.5%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:19:25.401Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗