GoApercu / Tapas.one
YC S26 · Application due 27 July 2026

VO2 Max, powered by SleepFM.
A Stanford-licensed digital biomarker.

Our YC S26 track is VO2 Max on the SleepFM foundation model — validated in Nature Medicine (January 2026) across 585,000+ hours of PSG data and ~65,000 participants, with 130 predictable conditions at C-index ≥ 0.75. SLAM / nano-cream is long-horizon R&D, routed to EIC Pathfinder and NIH SBIR.

🔒 Secure Download (Google SSO)Try the VO₂ estimator →Review the packagevo2.sleepfm.one →EU/Global/SBIR vision →

Founding team status is explicit: founder full-time; three technical leads part-time pending seed close; CTO offer outstanding. We state this plainly rather than imply otherwise.

The Missing Vital Sign

VO₂ max is the control signal for cardiometabolic disease.

HbA1c, blood pressure, BNP, BMI each measure a single organ. VO₂ max measures the integrated output of the whole system — heart, lungs, blood, vessels, muscle metabolism — and equals or beats those risk factors at predicting death. It is the only such marker estimable passively and continuously from a wearable already on hundreds of millions of wrists. AeroGlyphics ↗ is our public-facing evidence + estimator surface for this thesis.

44%
All-cause mortality drop, unfit → fit quartile (Mandsager, JAMA Netw Open 2018)
−14%
Cardiometabolic mortality per 1-MET gain, no ceiling (Kodama, JAMA 2009)
5.04×
Hazard ratio for lowest fitness quintile vs elite (n=122,007)
$413B
US diabetes cost/yr · AHA projects 61% of adults with CVD by 2050
Live estimator — feel the signal in 5s →SLAM methods →

Estimator baseline: VO₂ ≈ 15 × (HRmax / HRrest) · Uth-Sørensen-Overgaard-Pedersen; HRmax = 208 − 0.7 × age · Tanaka, JACC 2001. Norms via FRIEND registry (Kaminsky, Mayo Clin Proc 2015). Not a clinical device — smartwatch VO₂ or CPET is a valid input.

Three tracks, one platform

YC narrative leads with VO2 Max + the Singapore enterprise pilot. SLAM is footnoted as long-horizon R&D.

YC S26

VO2 Max — Stanford-licensed digital biomarker

Consumer + enterprise app built on the SleepFM foundation model (Nature Medicine, January 2026). Direct-to-consumer VO2 Max estimation from wrist wearables, with enterprise licensing to insurers and health systems.

Enterprise pilot

Singapore health-system pilot

Phased 500-member pilot (months 1–6) with a Singapore health-system / insurer partner. Reference deal for the platform business model behind VO2 Max — English-language market, Smart Nation data infrastructure, and MOH-aligned digital health regulation.

EIC Pathfinder + NIH SBIR

SLAM / HealthOS (long-horizon R&D)

4-Layer Transformer + PEDOT:PSS nano-cream biosensor stack. Routed to EIC Pathfinder and NIH SBIR Phase I — not part of the YC S26 narrative.

Compliance & data residency — Singapore pilot operates under PDPA (2012, amended 2020), aligns with MOH HCSA digital-health licensing, and IMDA's Trusted Data Sharing Framework. All PHI stored in-region on SG-hosted infrastructure; de-identified training data governed by a DPA.

Solutions

A digital definition of the biomarker — from wrist to plan to enterprise.

Life-science digital transformation, reforged for the wearable era. We bring a Bay Area foundation-model stack together with Boston clinical & regulatory rigor — MIT-trained engineering, Stanford-trained product, and an MD/PhD/MBA CMO. Not a consulting deck: a working platform your team can ship on.

Digital biomarker lifecycle
NO. 01
Digital Biomarker Profile
Define the digital blueprint of a biomarker — target claim, Critical Signal Attributes, sensor spec, reference standard (CPET, PSG, HRV), population norms, and intended use — before a single line of model code is written.
NO. 02
Signal & Cohort Genealogy
Automate lineage across wearable firmware versions, cohorts, and label sources. Visualize how a v0.9 Apple Watch cohort compares to a v1.2 Fitbit cohort — critical for FDA SaMD change-control.
NO. 03
Foundation-Model Execution (SleepFM-class)
Fine-tune Stanford-licensed foundation models (SleepFM, PSG-Foundation) on your data. Turn 585K+ hours of PSG and wrist-PPG into a production biomarker in weeks, not fine-tuning quarters.
NO. 04
Model Tech Transfer
Move a biomarker from a Cambridge/MIT lab notebook to a Bay Area production stack — TorchScript / CoreML / TFLite, on-device inference, MLflow lineage, and a signed model card as a first-class regulatory artifact.
NO. 05
Device & Facility Fit
Digitize the sensor-to-claim gap across Apple Watch, Pixel Watch, WHOOP, Oura, Garmin, and clinical-grade Holter. Instantly surface which SKUs meet the label and which need a firmware or sampling-rate change.
NO. 06
Commercial Launch Automation
Ship the biomarker into iOS/Android, HealthKit, Google Fit, and enterprise dashboards from one config. HIPAA + PDPA-ready pipelines, BAAs pre-wired on AWS us-east-1 and ap-southeast-1.
NO. 07
Digital Evidence Release
Automate reconciliation of real-world biomarker output against 510(k)/De Novo release specs. Every batch of predictions ships with an audit-grade evidence packet — the FDA-facing analogue of a batch release record.
Enterprise digitization
Electronic Evidence Records for R&D
Replace ad-hoc Jupyter + Slack workflows with signed electronic evidence records for early-phase biomarker development. Low-code drag-and-drop for clinical teams; pro-code hooks for ML engineers. GxP-adjacent from day one.
Electronic Evidence Records for Commercial
Automate cohort management, model monitoring, drift deviations, and post-market surveillance. Cut biomarker review cycles from weeks to days — the same operational lift EHR vendors achieved for clinical documentation.
Digital Claim Release Cockpit
A Boston-clinical / SV-product cockpit that reconciles market claims (VO₂ Max, sleep apnea risk, AFib burden) with live production data. Prioritize the reviews that actually gate revenue.
Services — SV product velocity, Boston clinical spine
Business Transformation
Wayne Pan MD/PhD/MBA leads payer- and health-system-facing transformation from Oakland. We translate a wearable signal into a covered benefit, a Star Rating lever, or an insurer-underwriting input — not a demo.
Technology & Data
MIT/Boston-trained engineering leadership on enterprise architecture, HL7/FHIR, HealthKit, Terra, Rook, and foundation-model MLOps. Data ontologies that survive an FDA pre-sub meeting.
Foundation-Model Services
Stanford-licensed SleepFM integration, custom PSG/PPG fine-tuning, and on-device distillation. Platform-agnostic across Apple, Google, Samsung, and clinical-grade device stacks — grounded in real wrist-floor data.
Delivery & Regulatory
Agile delivery with a clinical spine: pre-sub → 510(k) / De Novo → post-market. Voice of the clinician and voice of the patient, on time and on budget, with a CMO in the room.
Design Thinking for Clinical AI
Stanford-trained product leadership (Eangelica Aton, MD candidate) drives change management for clinicians, ops, and end-users. Beyond training — we design for the 3am nurse and the C-suite dashboard in the same sprint.

Front-End Study (FES) engagements available for insurers, health systems, and pharma digital-health teams — translate a wearable signal into a covered benefit, a claim, or a Star Rating lever in 6–8 weeks.

Team — as it actually stands

Three-founder core with a clinician, product/MD candidate, and CTO. Advisors round out modeling, mobile, nano, and governance.

Founders

Wayne Pan, MD, PhD, MBA
Chief Medical Officer — Oakland / SF Bay Area
Eangelica Aton
Chief Product Officer, MD candidate — Stanford-trained product & engineering
Michael Tedescucci
Chief Technology Officer — Boston area, long-term collaborator since 2021

Advisors

John Tillinghast, PhD
Advisor — stochastic modeling (SLAM). Founder status & equity: open loop
Elizabeth Mezias
Advisor — Mobile Architect (AI-assisted development)
Pau Sabater Nacher
Advisor — Nano lead (part-time, PhD-track)
Jan Vandenbos
Board of Advisors

Numbers

$500,000
YC pre-seed note
$27.5M
Year-5 ARR (base)
Jan 2026
Nature Medicine — SleepFM
$300–500K
NIH SBIR Phase I (SLAM)

$27.5M Year-5 ARR is the VO2 Max + enterprise-licensing base case. The $42M SLAM-inclusive upside sits in the appendix, not the headline. Pre-seed $1.5M supersedes the earlier $2.5M blended figure (which included SBIR non-dilutive).

YC + SBIR Package — v3

Updated 13 July 2026 — Singapore-first reframing (PDPA/HCSA compliance note, Singapore partner shortlist appendix, regenerated Executive Summary + Business Plan + Pitch Deck, YC single-slide addendum). China / Ping An references removed throughout.

★ Full YC Submission Bundle (PDF)
GoApercu_SLAM_YC_Submission.pdf
v2
Due Diligence Response v2 (Q1–Q4)
GoApercu_SLAM_DueDiligence_Response_v2.docx
v2
Pitch Deck (Business) — v3
GoApercu_SLAM_Business.pptx
v2
YC Slide Addendum (single-slide reframing)
GoApercu_SLAM_YC_Addendum.pptx
v2
Technical Deck
GoApercu_SLAM_Technical.pptx
v2
Executive Summary — v3
GoApercu_SLAM_Executive_Summary.docx
v2
Business Plan — v3 (Singapore-first)
GoApercu_SLAM_Business_Plan.docx
v2
Singapore Enterprise Pilot — Appendix
GoApercu_SLAM_Singapore_Appendix.docx
v2
Compliance FAQ (PDPA/HCSA/HIPAA/GDPR)
GoApercu_SLAM_Compliance_FAQ.docx
v2
Financials
GoApercu_SLAM_Financials.xlsx
v2
Architecture
GoApercu_SLAM_Architecture.docx
v2
Reference Diagram
GoApercu_SLAM_Reference_Diagram.docx
v2
IP Asset Checklist
GoApercu_SLAM_IP_Asset_Checklist.docx
v2

Compliance FAQ

PDPA · MOH HCSA · IMDA TDSF · HIPAA · GDPR — short answers; full detail in the Compliance FAQ doc.

Q
Where is user health data stored?
Singapore-pilot PHI is stored in-region on SG-hosted infrastructure (AWS ap-southeast-1). US consumer VO2 Max data is stored in us-east-1. No cross-border replication of raw PHI without an executed DPA.
Q
Singapore PDPA compliance?
GoApercu operates as a Data Intermediary under PDPA (2012, amended 2020). Explicit consent, right to withdraw/erasure, 72-hour PDPC breach notification, named DPO before pilot go-live.
Q
MOH HCSA licensing?
The Singapore pilot is structured to align with the Healthcare Services Act digital-health regime. VO2 Max is positioned as a wellness/decision-support biomarker; clinical-decision claims are gated behind HSA MDR registration (out of scope for the pilot).
Q
IMDA Trusted Data Sharing?
De-identified training data governed by bilateral DPA under IMDA's TDSF. Re-identification contractually prohibited. Partner receives aggregated cohort analytics via permissioned API — no bulk PHI export.
Q
HIPAA-ready for US enterprise?
Yes — HIPAA-eligible on AWS/GCP BAA services, AES-256 at rest, TLS 1.3 in transit, RBAC + audit logs. BAA executed before any US covered-entity pilot.
Q
GDPR / EU?
Not in scope for YC S26. SLAM long-horizon track routed to EIC Pathfinder will operate under GDPR with EU data residency and DPIAs per Article 35.