Therapeutics · Neuroscience and AI
Exin Therapeutics
Three Oxford neuroscientists building circuit level gene therapies for epilepsy, guided by AI models trained on mouse brain recordings. In late 2025 they moved the wet lab out of San Francisco and into Manila to make that data cheaper.
A post on a Saturday
On November 1, 2025, Marko Tvrdic put a short note on LinkedIn. “We are moving our wet-lab operations from San Francisco to Manila, Philippines, where we’re building a neuroscience data factory, a new model for large-scale brain data generation.”
A data factory, in the company’s telling, is a room of automated rigs running mouse neurophysiology around the clock, with version controlled pipelines from raw signal to feature extraction. The output is not a paper. It is validated data points per dollar, and the company blog reduces the plan to that ratio: cheaper rent, cheaper skilled labour, more rigs, more data. “Achieving this level of scale is impossible within the artisanal workflows of today’s academic and biotech labs,” the post says.
Two weeks earlier the Philippine Board of Investments had made the move official. On October 17, 2025 it announced a Green Lane Certificate under Executive Order No. 18, the fast track for strategic investments, covering a neurotherapeutics laboratory in Bonifacio Global City, Taguig. Phase one: 45 million pesos, 2025 to 2027, 35 jobs. Phase two: a billion pesos in Clark, Pampanga, 2027 to 2029. The BOI called it the first hub of its kind in Southeast Asia.
Why brain drugs keep failing
Exin’s starting claim is a familiar one, and the company states it plainly. Many neurological and neuropsychiatric symptoms come from an excitatory and inhibitory imbalance in cortical circuits. “Think of it like an orchestra where some instruments are playing too loud while others are too quiet,” the launch post says. In epilepsy the overactive players cause seizures. In autism the same imbalance can present as sensory hypersensitivity.
The therapeutic bet follows. “Instead of trying to restore missing genes, we’re supplementing already existing genes that we know will be directly changing the neural activity in the brain,” the company says. That is a different proposition from most gene therapy, which aims at one defective gene in one named disease. Exin aims at the tuning of a circuit, which in principle does not care which mutation put it out of tune. The harder problem is measurement, and Exin argues the translation gap is less about bad animal models than about not knowing which of their readouts predict anything in humans. That is the case for collecting more data than a normal lab can afford. At this stage it is a case rather than a result.
Three from one Oxford building
Gabriel Ocana Santero has a pharmacology PhD from Oxford and was a Marie Sklodowska Curie fellow. Ivan Alex Priela Lazarte trained as a physicist in the Philippines and Taiwan before an Oxford neuroscience PhD as a Clarendon Scholar. Marko Tvrdic is on leave from his Oxford PhD.
The public record confirms the lineage rather than taking it on trust. Ocana Santero is first author on a June 2025 Nature Communications paper on perinatal serotonin signalling and cortical GABAergic development; Lazarte is a co-author on it. Tvrdic is a co-author on a December 2025 Neuron paper. Both come out of the Butt and Packer labs at Oxford. Adam Packer, now at Meta, sits on Exin’s advisory board with Tom Otis of UCL, Alan Mardinly of Science Corporation and formerly Neuralink, Adam Ezra Cohen of Harvard, Mackenzie Mathis of EPFL, and Carlos Portera-Cailliau of UCLA, listed as unpaid. Lazarte’s brother has autism spectrum disorder, which the company names as his reason for the work. He is also Filipino, which is the thread running into the rest of this story.
Timeline
The first sale under the company’s Regulation D offering was November 18, 2024. California registration followed on January 10, 2025. The public launch came on January 13, 2025, in Y Combinator’s Winter 2025 batch. The YC profile describes the first six weeks in San Francisco: an animal lab opened in South San Francisco, a proof of concept in mice, a multimodal model, two provisional patent filings, “all while utilizing less than 50% of our YC funding.”
Then the Green Lane Certificate in September and October 2025, the Manila move in November, the Form D on February 10, 2026, and on May 13, 2026 a poster at the American Society of Gene and Cell Therapy meeting in Boston, where Lazarte would “present our progress over the past year in developing circuit-centric gene therapies, beginning with a programme focused on generalized epilepsy.”
What is proven, and what is still claimed
| Evidence | What the record shows | Source type |
|---|---|---|
| SEC Form D | Filed Feb 10, 2026, CIK 0002109179, Rule 506(b). First sale Nov 18, 2024. Offered and sold $2,162,300, nothing remaining, 28 investors. Securities marked as an option to acquire and a security to be acquired. Related persons: the three founders. | Public record |
| Corporate registration | Delaware corporation registered in California on Jan 10, 2025, document 6530638, active. Ocana Santero CEO and director, Lazarte CFO and director, Tvrdic secretary. | Public record |
| Philippine certification | Green Lane Certificate awarded Sep 17, 2025. Phase one 45 million pesos in Bonifacio Global City, 35 jobs, operations from October 2025. Phase two 1 billion pesos in Clark, 2027 to 2029. | Government record |
| Founders’ published work | Ocana-Santero G with Lazarte IP, Nature Communications, Jun 2025. Tvrdic M, Neuron, Dec 2025. Ocana-Santero G, ACS Chemical Neuroscience, Feb 2024, and Cerebral Cortex, Mar 2023. Adam Packer, an Exin advisor, co-authors three of the four. | Public record |
| Work published as Exin | A PubMed affiliation search returned nothing as of Sep 22, 2026. A PatSnap Synapse record carried no pipeline, patent or trial data. The two provisional patents claimed in 2025 would not publish. | Not found |
| Preclinical results | A proof of concept in mice, a model that directs screening, and a plan to validate 10 candidates a month in Manila. None published or independently reviewed. | Company-stated |
| Reported funding | Tracxn and Startup Intros both report $500,000 in one seed round, naming Y Combinator and Comal Ventures. The Form D reports $2,162,300 from 28 investors. | Differs from the filing |
| What the company sells | The 2025 launch described circuit level genetic therapies. The website in September 2026 describes a target discovery platform that is “Agnostic to therapeutic modality.” The May 2026 conference post still says gene therapies. | Company-stated |
Read plainly: the corporate and academic record is checkable, and the Philippine facility is documented by a government agency, not just the company. The science is not. No published data, no patent on file, no trial, no partner. What Exin has built so far is a cost structure and a data pipeline, plus biological claims resting on its own account of its own mice.
What to watch
- Whether the epilepsy program yields a disclosed candidate and an IND enabling package, or stays a poster.
- Whether phase two in Clark, Pampanga starts in 2027 as certified, and at what scale.
- A paper or preprint with Exin as the affiliation, which would move the platform claims out of the company-stated column.
- A new Form D for a priced round. The February 2026 filing reports convertible instruments, not equity.
- Whether either provisional filing becomes a published application about 18 months on.
In their words
“Think of it like an orchestra where some instruments are playing too loud while others are too quiet.”
Exin launch post, Y Combinator, January 2025 · Company
“Instead of trying to restore missing genes, we’re supplementing already existing genes that we know will be directly changing the neural activity in the brain”
Exin launch post, Y Combinator, January 2025 · Company
“all while utilizing less than 50% of our YC funding.”
Exin company profile, Y Combinator, 2025 · Company
“By lowering the cost per datapoint by an order of magnitude, we can generate the scale of multimodal datasets needed for predictive neurotherapeutic discovery.”
Marko Tvrdic, co-founder and COO, LinkedIn, November 1, 2025 · Company
“The Philippines offers a unique combination of scientific talent, cost efficiency, and a strong innovation environment”
Ivan Lazarte, co-founder and CSO, quoted by THEPHILBIZNEWS, October 21, 2025 · Independent
“It’s not just about attracting foreign investment, it’s about transferring cutting-edge expertise to Filipino scientists, building local capacity, and positioning the Philippines as a leader in advanced biotechnology research”
Ceferino Rodolfo, Trade Undersecretary and BOI Managing Head, October 17, 2025 · Government release
Related companies
Sources
- Public recordForm D, Exin Therapeutics, Inc., CIK 0002109179
- Public recordCalifornia Secretary of State registration, document 6530638
- Government recordBOI fast-tracks first neurotherapeutics lab in the PH
- Public recordPerinatal serotonin signalling and cortical GABAergic development
- Public recordRepresentations of economic variables in the frontal cortex
- Public recordSensory-evoked serotonin release in adult neocortex
- Public recordTemporal origin of mouse claustrum
- Not foundPubMed affiliation search for Exin Therapeutics
- Not foundPatSnap Synapse record, no pipeline or patent data
- IndependentAI biotech Exin Therapeutics plans 1-billion-peso expansion
- IndependentUS-based biotech firm cues P1.45-B investment in the Philippines
- IndependentUS neurotherapeutics lab to operate in PH
- IndependentPH bags first biotech investment
- IndependentExin Therapeutics profile and employee count
- IndependentExin Therapeutics: Funding, Team & Investors
- CompanyLowering the cost of data, raising the scale of discovery
- CompanyExin Therapeutics company page and launch post
- CompanyExin Therapeutics website, team and news pages
- CompanyASGCT 2026 poster note and the Manila move post
Profile researched and written by Healthcare Discovery. Last updated September 29, 2026.
