Company profile · 4 min read
Exin Therapeutics: multimodal discovery for neurotherapeutics
Exin Therapeutics uses multimodal neural data, AI and high-throughput mouse studies to prioritize neurotherapeutic candidates.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Exin Therapeutics is an early neurotherapeutics company using multimodal data and AI to prioritize drug candidates that change neural activity. Its likely partner is a pharma or biotech team that needs better preclinical signal before spending more time and money on a candidate—not a patient looking for a treatment today.
What it does
The YC profile describes AI models trained on high-density mouse data and an initial focus on epilepsy associated with autism and Parkinson’s disease. The company’s science page says the platform is modality-agnostic and aims to predict a candidate’s potential across neurological disorders by combining neural activity, behavior and transcriptomics.
The YC launch explains the workflow: use AI to identify disease-relevant circuit drivers, prioritize which therapies to test and provide efficacy readouts during internal experiments. The company’s public profile also reports a mouse proof of concept, an operating animal lab and two provisional patent submissions; those are company-reported development milestones, not clinical evidence.
Why I’d look closer
The scientific thesis is specific: target neural-circuit dysfunction rather than start with a single genetic background or one modality. The founders’ public context is relevant. Gabriel Ocana Santero is described as an Oxford-trained pharmacologist and neuroscientist with gene-therapy and NeuroAI work; Ivan Lazarte is a physicist and Oxford-trained neuroscientist; Marko Tvrdic has worked across iPSC, retinal, circadian and systems neuroscience.
The tradeoff is translation. A high-dimensional mouse signal can improve preclinical prioritization without predicting safety, dosing, human efficacy or regulatory success. The company is still a discovery and preclinical platform, and public pages do not publish a clinical candidate, trial status or licensing terms.
What I’d ask
Which neural-activity and behavioral endpoints are validated against independent experiments? How does the model avoid learning lab-specific artifacts? What is the modality, target and indication furthest along? Which data, patents and development rights would a pharma partner receive, and what remains to reach IND-enabling work?
My editorial take
Shortlist Exin for a preclinical partnership conversation if neurological discovery is strategic and you can evaluate the model against owned assays. Do not treat the AI score or mouse proof of concept as medical evidence. The company’s value will be decided by reproducible candidate prioritization and a credible translation path, not the novelty of the model alone.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | AI and high-throughput mouse platform for neurotherapeutic discovery |
| Buyer | Pharma, biotech and preclinical neuroscience partners |
| Focus named | Epilepsy associated with autism, Parkinson’s and neural-circuit disorders |
| Stage visible publicly | Discovery/preclinical; no clinical candidate or trial status published |
| Main question | Does multimodal neural data improve candidate selection in reproducible assays? |
Sources checked
| Source | Checked |
|---|---|
| YC company profile | 2026-09-19 |
| Exin science page | 2026-09-19 |
| Exin team page | 2026-09-19 |
| Exin YC launch | 2026-09-19 |
Cohort context
Exin Therapeutics is listed in Winter 2025. In our 2026-09-18 directory snapshot, 11 of 165 listed companies in that cohort have YC’s primary industry label Healthcare (6.7%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:19:33.624Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Not observed in this response |
| Canonical link | Not observed in this response |
| H1 or H2 heading | Observed |
| Typed structured data | Not observed in this response |
| 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 ↗