Clinical Research Scientist, Precision Psychiatry

About Psyrin

Psyrin builds AI for behavioral health. Our voice agent, Mira, runs the clinical interviews around a behavioral health visit: intake, consent, screening, and a structured psychiatric pre-assessment. It then drafts the assessment and treatment plan documentation for the clinician to review and sign.

Across our partner clinics, Mira has given clinicians 3.3× the assessment capacity per hour of manual assessment (as of September 2026). Psyrin is live with multiple behavioral health organizations, most of them community mental health and addiction providers, as well as work with state government. Our work serves Medicare and Medicaid patients, some of the most vulnerable populations in America.

The company was founded by doctors, psychologists, and AI researchers, largely from King's College London. We're published in Nature and backed by top thesis-driven investors.

While others build surface-level apps, we dig deep into the root causes of psychiatry's problems: inconsistent assessment quality, subjective diagnoses, and fragmented measurement systems. By capturing not just what patients say but how they say it, we're working toward our long-term vision: democratizing access to psychiatric evaluation and building precision psychiatry with AI. In short, an AI psychiatrist, built with the scientific rigor that mental healthcare demands.

The future of psychiatry

Most of medicine has objective tests. A cardiologist reads an ECG; an oncologist sequences a tumor to choose the drug most likely to work. Psychiatry still relies almost entirely on conversation, rating scales, and one clinician's judgment, often in a single short appointment. That is about to change. Speech, facial expression, movement, sleep, and heart rate all carry signals of how someone is doing, and for the first time we can measure them at scale and over time.

Precision psychiatry will mean matching each patient to the treatment most likely to work for them, the first time. Today, finding the right medication or therapy is often trial and error: months on one option, then the next. Measurement will change that. It will mean predicting who will respond to an antidepressant, TMS, or esketamine before they start, spotting relapse before it becomes a crisis, and following recovery between visits instead of guessing at it from one appointment. It will ground diagnosis in measurement too, telling apart conditions that look alike but need different care, such as bipolar and unipolar depression. And it will bring expert-level psychiatric care to people who today wait months to see anyone at all.

We've laid the groundwork. Psyrin's patent-pending voice biomarker research has been published in Translational Psychiatry (Nature Portfolio), and covers response to TMS and esketamine, relapse forecasting (PRIME), psychosis prediction, schizotypy, and differential diagnosis. Our consented clinical dataset grows every week across our partner clinics.

The studies you run here will decide which of these signals hold up in real clinics, with real patients, and become measures clinicians use every day. If you want your research to leave the journal and change how psychiatry is practiced, this is the place to do it.

The role

You'll be Psyrin's first dedicated research scientist and a founding member of Psyrin Superpsychiatry Labs. Led by the CTO, the lab brings together clinical research, ML research, and AI engineering, working closely with the COO. Its goal is psychiatric assessment that is more objective, longitudinal, and scalable than any single clinician can deliver. You'll own the clinical science behind that goal. Psyrin's founders built a voice biomarker program. Your job is to extend it into a multimodal precision psychiatry program that is rigorous, reproducible, and ready for regulators. Matching patients to the treatments most likely to help them comes first, with more objective diagnosis alongside it. That program brings together speech, language, facial expression and movement from video, sleep, activity, and heart rate data from wearables, and structured clinical and EHR data.

That means designing prospective studies at US partner clinics and getting them through IRB review. You'll analyze multimodal data with the engineering team, and publish in peer-reviewed journals. You'll publish your work under your own name, contribute research to Psyrin's regulatory program, fund the work through grants, and run partnerships with academic groups. You'll report to the CTO and work closely with the COO, Psyrin's clinicians, and the clinic leaders whose patients make the research possible.

What you'll do

Run the precision psychiatry research program

Set and maintain the research agenda with the CTO, with treatment selection and response as the priority. Decide which treatments, conditions, modalities, and populations come next, judged by what moves Psyrin closer to an AI psychiatrist.
Lead treatment-matching studies: predict who will respond to which medication, TMS, esketamine, or psychotherapy, and detect early response or non-response within weeks of starting.
Build measures that make diagnosis more objective, especially for conditions that look alike early on but need different treatment, such as bipolar and unipolar depression, or emerging psychosis.
Build a multimodal biomarker program with voice at its core:
- Voice (our key modality): acoustic, prosodic, and linguistic markers from Mira sessions, building on Psyrin's published and patent-pending work on treatment response, relapse forecasting, psychosis, schizotypy, and differential diagnosis.
- Video: facial affect, psychomotor signs, and eye gaze from telehealth and kiosk sessions.
- Wearables: sleep, activity, and heart rate variability from Apple HealthKit, Android Health Connect, and research-grade devices.
- Clinical data: prescribing and outcome data, EHR history, and validated scales (PHQ-9, GAD-7, C-SSRS).
Design prospective and retrospective studies: protocols, power calculations, outcome definitions (e.g. response and remission on validated scales), diagnostic reference standards (e.g. SCID-5, clinician final diagnosis), and pre-registered analysis plans. Include longitudinal designs that follow patients from the start of treatment through response, remission, or relapse.
Design studies that test whether multimodal measures predict treatment response better than any single one, or than current practice, with honest baselines.
Analyze multimodal and longitudinal data with the engineering and ML team. Own the statistics, validation design, bias and subgroup analysis, and reporting standards (TRIPOD+AI, STARD).
Publish as first or senior author: lead manuscripts from first draft to acceptance, and present at conferences.

Own IRB, ethics, and data governance

Write and submit protocols to central and local IRBs. Manage amendments, continuing review, and adverse event reporting.
Design consent processes, with partner sites, for voice, video, and wearable data collected in clinic, by phone, and at home.
Make sure research data handling meets HIPAA and, for UK collaborations, UK GDPR. Define de-identification, data access, and retention for the research dataset.

Contribute to regulatory evidence

Psyrin leads FDA strategy and submissions. You'll design and run the studies that feed them.
Design analytical and clinical validation studies to the standard regulators expect, and keep study documentation inspection-ready.
Advise Psyrin's FDA work on what the research can and can't yet support.

Win grants

Find and write funding applications, including NIH/NIMH SBIR and STTR, and NIHR or Innovate UK for UK-led work.
Act as PI or co-investigator, and manage reporting for awarded grants.

Run academic partnerships

Manage existing collaborations, including with King's College London, Imperial College London, and Yale University, and build new ones with US academic medical centers.
Agree data-sharing, authorship, and IP terms with partners and Psyrin's legal advisors.

What we're looking for

A PhD in psychiatry, clinical psychology, psychiatric epidemiology, neuroscience, biomedical engineering, computer science, or a related field (or an MD or equivalent, e.g. MBBS, with substantial research training).
2+ years of post-doctoral, academic clinical, or industry research on human-subject clinical studies.
A first- or senior-author record in peer-reviewed journals.
Hands-on experience writing IRB protocols and running studies with clinical populations, ideally treatment or outcome studies in people with serious mental illness or substance use disorders.
Strong statistics: prediction modeling and diagnostic accuracy (sensitivity, specificity, AUC, calibration), longitudinal and mixed-effects models, treatment-effect heterogeneity, and handling missing and imbalanced data.
Experience with at least one kind of digital or physiological data, such as speech, video, wearable or sensor data, or large EHR datasets, plus the curiosity to work across all of them.
Fluency in Python or R, and comfort working with data and ML engineers on model development and validation.
Clear writing for three audiences: journal reviewers, grant panels, and clinic medical directors.
Willingness to complete HIPAA training and work under Psyrin's data protection policies.
Comfort working in an early-stage company, where you'll set up processes as well as follow them.

Nice to have

Experience with treatment-response prediction or clinical trials of psychiatric treatments, such as medication, TMS, esketamine, or psychotherapy.
Research on digital phenotyping or multimodal biomarkers: acoustic and prosodic features, NLP, computer vision for affect or movement, or passive sensing from phones and wearables.
Experience with multimodal or foundation models (speech, vision, or time series) and with fusing modalities.
Evidence generation for an FDA-regulated digital health product, or experience with SaMD, IEC 62304, or ISO 14155.
An awarded NIH, SBIR/STTR, NIHR, or similar grant as PI or co-I.
Experience in community behavioral health, addiction services, or psychosis early-intervention settings.
For PhD candidates, clinical training or licensure (e.g. PsyD, DClinPsy, LCSW). For MD candidates, formal research training (e.g. MSc, MPH, or a research fellowship).
An existing network in US academic psychiatry.

Location and travel

Based in the San Francisco Bay Area.
Expect about 10% travel to US partner clinics, academic collaborators, and conferences.
We meet in person for a team retreat once a quarter.
Collaboration with London-based colleagues means some early-morning calls.
You must already have the right to work in the US, for example as a citizen, a green card holder, or on a valid work visa. We'll consider O-1 sponsorship for exceptional candidates only.

Compensation and benefits

A competitive base salary, depending on experience and location.
A meaningful equity grant.
Benefits: unlimited PTO and quarterly in-person team retreats, plus 100% of premiums paid for PPO Platinum (or equivalent) health insurance plus dental and vision, a safe harbor 401(k) with a 3–5% employer match, one round-trip flight home a year, and Groundfloor membership for Bay Area staff. Full details on our careers page.
Conference attendance and open-access publication fees covered.
All work travel costs covered.

How to apply

Send your CV and publication list to careers@psyrin.ai with a short note on one clinical study you designed or ran: what went wrong, and what you did about it.

Psyrin is an equal opportunity employer. We welcome applicants of every background and don't discriminate on the basis of race, religion, color, national origin, sex, sexual orientation, gender identity, age, disability, veteran status, or any other protected characteristic.