Investor deck · Seed · 2026

Supervised AI in mental health

Guided voice support for continuity of care

ReMind 01 / The problem

The problem

Therapy happens in sessions.
The work happens between them.

Patients leave the therapy room with a question to sit with, and are alone with it until the next appointment. Clinicians arrive at each session reconstructing the week they couldn't see.

The current AI mental health market answers this with consumer chatbots attempting to substitute for clinicians. Regulators, clinical bodies, and patients themselves are growing cautious of unsupervised AI in mental health.

The clinical relationship is the active ingredient. The AI should be in service to it.

ReMind 02 / What ReMind does

What ReMind does

A voice support tool that lives in the space between sessions, supervised by the patient's clinician.

The patient speaks a short check-in in the morning. ReMind returns a guided audio reflection later in the day. Every reflection is shaped by a guidance profile authored by the clinician. No advice. No behavioural suggestions. Just continuity.

FIG. 01: A DAY, BETWEEN SESSIONS Time →

ReMind 03 / How it works

How it works

Every reflection is composed from three inputs.

→ INPUT 01

The guidance profile

The clinician's encoded voice and method. Authored by the clinician or assigned from a library of profiles authored by experienced clinicians for particular presentations.

→ INPUT 02

The focus

The current therapeutic target, set with the patient in session.

→ INPUT 03

The recording

What the patient just voiced.

The composition is the IP

The composition is what defines ReMind. The guidance profile is where the clinician's craft is encoded.

ReMind's data architecture is privacy-by-design: patient content stays on the patient's device, shared recordings are end-to-end encrypted and deleted within 72 hours.

ReMind 04 / The pilot

The pilot

The architecture works at small scale.

Three-week feasibility and acceptability study with 27 participants on Prolific, screened for moderate self-reported anxiety in a workplace context. The full supervised architecture deployed end-to-end, with a guidance profile authored by Edward Sim and a named guide on the receiving end.

82.22/100

SUS · System Usability Scale

Excellent. Top quartile of digital health.

4.04/5.00

uMARS · App quality

Above published benchmarks for mental health apps.

7.15/10

Between-session use intent

70% of participants scored 7 or higher.

"The reflections felt personalised but unbiased, like an outsider not judging you but providing insight into your patterns."Pilot participant

"I treated it like a companion. It gave me the confidence to move forward."Pilot participant

The mechanism we'll scale is the one the pilot tested.

ReMind 05 / Go-to-market & platform

Go-to-market & platform

Wedge into partners. Platform underneath.

Reach patients through the partners who already serve them. Build one technical foundation that supports multiple commercial surfaces.

→ WEDGE

Through marketplaces & clinics

One-year free deployments with therapy marketplaces and clinics, in exchange for usability and outcomes data.

Convert to paid commercial relationships, priced per-seat or per-patient, with ReMind positioned as a benefit offered through the partner.

Active conversations · UK clinical networks

→ PLATFORM

One architecture, many surfaces

ReMind Voice is the first variant of a platform that composes guidance profile + focus + recording into voice reflections.

A second product surface is in development: research instrumentation for longitudinal experience capture in clinical trials.

Active conversations · psychedelic research investors

→ WHY NOW

The supervised wave is arriving

Regulatory pressure on autonomous AI in mental health is intensifying. Therapist supply is constrained. Patients are signalling demand for between-session support.

The companies positioned for the supervised wave will define the category.

Market timing · 2026

ReMind 06 / Team

Team

Technical depth in voice AI. Clinical depth in digital therapeutics.

Founder

Ray McClure

Founder. Product architect and creative technologist with a career building products around emerging interaction patterns. Founding team member at Twitter. Creator of the Stop Motion effect at TikTok, now used in over 250 million posts. Led Dreamboat Studio for a decade, building experiential work for Google, The New Yorker, and Vogue. VVVR (shared voice-controlled VR) received an Epic MegaGrant, premiered at David Lynch's Festival of Disruption, and features in Peter Rubin's Future Presence and the Stanford d.school. Built the ReMind platform end to end: voice AI pipeline, guidance profile architecture, iOS application, and backend. UK Global Talent Visa.

Co-founder

Jeremy LaTrasse

Co-founder Twitter, Co-founder Message Bus, Heavy Bit, Square, Sightglass Coffee.

Clinical Advisor

Heather Bolton

Clinical Psychologist; ex-Unmind Science Lead. Led global Science function through Series C growth and AI pivot at Unmind. Built and scaled digital therapeutic services across 40+ countries. Published RCTs and high-impact research.

Clinical Contributor

Edward Sim

Cognitive Behavioural Psychotherapist and founder of Ed can Help. Provided ongoing input on guidance profile creation, including the authoring flow and the content of the profile used in the pilot.

ReMind 07 / Traction & the ask

Traction & the ask

Raising £400k. 12–18 months of runway.

Pilot complete · platform abstraction near complete · partner-facing site live

Where we are

Mechanism validated, partners in motion.

Pilot completed with strong results
Platform abstraction near complete
UK clinical networks in active conversation
Psychedelic research investors in active conversation

Use of funds

Deploy, study, and build the regulatory path.

Real-world deployment via marketplace and clinic partners
Follow-on clinical study with academic collaboration
Class IIa groundwork (MHRA pre-sub, IDAP, QMS)
Platform & product development, team

Commercial structure

B2B2C: free in year one, licensed thereafter.

Year-one deployment provided at no cost in exchange for usability and outcomes data
Per-patient or per-seat licensing with partner organisations post-deployment
Partner organisation is the buyer; the platform sits alongside their existing clinical work
Platform extensibility supports a parallel research instrumentation line

The path

Seed earns the right. Series A funds the proof.

Seed earns the right to pursue Class IIa at Series A. Series A funds the formal clinical investigation and conformity assessment, with the regulatory path de-risked by the seed period work.

The trajectory

Seed → real-world deployments & outcomes data → Series A → Class IIa clinical investigation.

ReMind 08 / Why now, why us

Why now, why us

The market is moving from autonomous to supervised AI. ReMind is built for it.

→ 01

Technical depth in voice AI from a founder who's built platform-scale products before.

→ 02

Clinical depth from advisors who've built and scaled digital therapeutics to commercial success.

→ 03

A pilot that validated the supervised architecture end-to-end.

→ 04

A platform that supports multiple commercial surfaces from a single technical foundation.

→ 05

A go-to-market designed around the partners who already serve the patients we want to reach.