Always-on precision health.
Your body, continuously decoded.
An always-on intelligence layer that turns episodic care into continuous, MD-signed precision health across genome, biomarkers, wearables, and behavior.
One stack. Sold underneath every clinic on the network. The platform compounds with multi-tenant data, the operating arm proves it works, and the regulators want it tomorrow.
Specialty pharma platforms, healthcare-services rollups, and clinic networks all face the same wall: clinical decision quality scales with clinician hours, and compliance overhead caps how fast any AI tool can get deployed.
The AI tools that would help are stuck behind BAA reviews, cross-border data-residency carve-outs, and per-token billing that scales the wrong way.
Not a consumer brand. We sell the substrate that lives underneath these four buyer shapes:
The platform layer every clinic on the network shares:
Above 31 billion parameters, open clinical models stopped hallucinating — Bo Wang's UHN AI Hub benchmark, May 2026, Gemma 4 31B at 86.5% diagnostic accuracy with zero hallucinations on diagnostic tasks.
That single benchmark is what makes on-device clinical AI viable today. A year ago it was a research bet. A year from now it's a competitive cost. We're inside the window where moving first matters.
Consumer telehealth captures one slice of the patient. Single-clinic concierge captures one location. Pharma reps capture one product line. Nobody has built the audit-grade clinical decision layer underneath all of them. That layer is what we sell.
Every clinic on the network sharpens it. The competitor with no installed base cannot catch up — they have no decision history to learn from.
Consumer brand + DTC fulfilment
Hims & Hers · Ro · Eucalyptus
Annual biomarker labs
Function · Protocol Health · MORROW
Coaching app + wearable
Healthify · Ultrahuman · Twin Health
Pharmacy distribution
Truemeds · 1mg · Apollo 24/7
Single-location clinic concierge
Biopeak · Sukino · SG saturated set
Gray-channel peptide aesthetic
ARQ · TAC · Mumbai chains
The audit-grade clinical decision + monitoring + dispensing platform underneath all of them — is unbuilt.
Pricing pressure on each slice keeps competitors inside their slice. Building the cross-slice substrate requires regulatory chops + AI infra + capital patience to sit beneath the layer that earns. We're the team where those three meet.
Our technical partnership has shipped on-premise AI deployments at the BSE (Bombay Stock Exchange) / NSE (National Stock Exchange) / top-15 Indian banks — the same compliance-by-construction architecture Omnis uses for HIPAA (US health-privacy law) / DPDP (India's 2023 data-protection act) / PDPA (Singapore + Thailand). Not a research lab. A production line.
The gap is structural. The technical window opened six months ago (Bo Wang UHN benchmark, Gemma 4 31B). The shape of buyer who needs this layer is already operating but missing the AI overlaythat lets their clinical quality scale faster than headcount. We're the layer they all need next.
The window is wide enough to walk through. It is not wide enough to wait through.
The platform. Sold B2B to clinic networks, hospital systems, telehealth operators. Audit-grade prescribing decisions, continuous watcher infrastructure, on-device clinical AI, full provenance trail.
The operating arm. Two operators. The clinical front-end in select markets — telemedicine + specialist physician panel + partnered dispensing + fulfillment. The credentialed proof point for the platform underneath.
Omnis Digital Clinic runs the front-end — patients arrive, intake runs, the gate pipeline pre-drafts a protocol, the specialist MD signs off, the protocol dispenses through a partner pharmacy, the watcher infrastructure monitors continuously. Every one of those events writes to the Omnis platform. Meanwhile a hospital network in another geography licenses Omnis underneath their existing clinical operations — zero patients shared, multi-tenant data isolated, but the gate-pipeline weights and outcome benchmarks they get back are richer because the clinic is feeding them too.
Two layers. Same engineering team. One investable thesis.
Every patient starts with a structured intake: bloods (HbA1c [3-month average blood sugar], ApoB [a cardiovascular-risk lipoprotein], ALT [liver enzyme], eGFR [kidney filtration], TSH [thyroid hormone], vitamin D — extendable per archetype), wearable sync (Whoop / Oura / Apple Health), full medication and supplement list, family history, current life context.
Plus archetype routing: the patient profile triages into one of five operating archetypes — Founder Optimizer, Aesthetic Mid-Lifer, Perimenopausal Reset, Metabolic Recovery Returner, Diaspora Returner. Archetype determines content layer, monitoring cadence, and acquisition channel.
The AI drafts a protocol against the patient's full record. Every line item — drug class, dose, titration schedule, adjunct supplements, monitoring cadence — passes through eight explicit gates: safety, efficacy, adherence, cost, drug interactions, timing, contraindications, dosing precision.
Each gate logs the evidence it consulted (which guideline edition, which lab values + timestamps, which interaction database). The output isn't just a protocol — it's a protocol plus a reproducible reasoning chain.
The drafted protocol queues to a specialist physician (endocrinologist / internal-medicine specialist / cardiologist per DCGI [Drugs Controller General of India] 24 March 2026 advisory for GLP-1 [the Ozempic / Wegovy drug class]; cardiology / metabolic / aesthetic specialty per indication). The MD (medical doctor) sees the AI's reasoning, the gate-by-gate evidence, and can approve, edit, or reject.
Every MD edit becomes training signal — the rejection reason, the dose adjustment, the alternative compound preference all flow back into the gate-pipeline weights. The platform sharpens with every clinician interaction.
Signed protocol routes to a Form-20 partnered pharmacy in the patient's revenue district (or international shipping for cross-jurisdiction patients). CDSCO (Central Drugs Standard Control Organisation — India's national drug regulator) approved compounds only — branded generics, approved biologics, no gray-market peptides. Cold chain managed. Patient receives + onboards.
Revenue model: rev share on fulfilled product flows back to Omnis. Omnis Digital Clinic handles fulfilment in-house at our operated sites; licensed partner clinics retain their existing pharmacy partnerships.
Patient enters the standing monitoring loop. Guideline watcher (CPIC [pharmacogenomic dosing], ADA, FDA labels, EMA SmPC [the EU drug-label format]) re-runs gates when relevant guidelines update. Trial watcher (ClinicalTrials.gov, CTRI [India's trial registry]) surfaces enrollment matches. Literature watcher (the major medical journals — NEJM, Lancet, plus preprints on bioRxiv) flags new evidence affecting current stack.
All of this runs locally on the clinic's on-device appliance — Gemma 4 31B or Qwen3.5-35B, 86.5% diagnostic accuracy per Bo Wang's May 2026 UHN benchmark. PHI never crosses the building's edge. HIPAA / PDPA / DPDP compliance by architecture.
One canonical patient record. Five stages of decisions, each logged with full evidence provenance. Standing watcher infrastructure that re-evaluates every record when a guideline updates, a trial opens, or new literature lands. On-device clinical AI that keeps PHI inside the building. The architectural shape is audit-grade decisions with continuous improvement. The clinical scope — obesity, metabolic, longevity, peptide-adjacent, oncology adjuncts — is configurable per partner clinic.
Cross-checks allergies, organ dysfunction (eGFR, LFTs, ECG if relevant), drug-class history, and black-box warnings.
Predicts expected response for this patient's phenotype using RCT subgroup data + real-world cohort evidence.
Predicts whether the patient will actually take it — Rx fill history, injection-vs-oral preference, side-effect tolerance.
Net out-of-pocket calculation, formulary lookup, generic alternatives, prior-auth requirements.
Cross-references the full med + supplement list against interaction databases at the PK and PD level.
Is now the right cycle? Scheduled procedures, pregnancy plans, seasonal context, life events.
Absolute and relative contraindications — pregnancy, MTC history, MEN-2, pancreatitis history, gastroparesis.
Personalized starting dose, titration schedule, max dose, organ-function adjustments.
This is the moat made concrete. Two patients walk in with the same biomarkers. The gate pipeline reads their full context — adherence history, insurance, family history, current meds — and routes them to different protocols. The clinician sees both reasoning chains. Audit-grade by default.
→Same input labs, same recommended candidate class. Gate 03 (adherence), Gate 04 (cost), and Gate 07 (relative contraindication) flagged differently for Patient B — and every flag is reproducible from the same provenance log the auditor reads.
CPIC publishes a new dosing recommendation for a compound in this patient's stack. The watcher re-runs interpretation, drafts the adjustment, queues it for the specialist's sign-off. Median latency: hours, not 18 months.
Patient is on tirzepatide + has T2D + carries the metabolic-phenotype profile that ClinicalTrials.gov says a Phase 3 just opened for this week. Surfaced in tomorrow's clinical brief.
A bioRxiv paper drops on the dose-response interaction between two compounds on the patient's protocol. The literature watcher tags it, scores evidence tier, flags it to the clinician with the relevant section pre-extracted.
No clinic builds this. They can't — it requires standing engineering infrastructure running 24/7 across thousands of feeds. They subscribe to it. The watcher trio is the kind of overhead that only makes sense at platform scale, sold across many clinics. The clinic that licenses Omnis is the only clinic in their geography whose patients are getting guideline-current care without a 24-month lag.
Above 31 billion parameters, open clinical-AI models stopped hallucinating.
Bo Wang's UHN AI Hub benchmark (May 2026) tested 188 models on general disease diagnosis. Gemma 4 31B hit 86.5% accuracy. Fine-tuned Qwen3.5-35B hit ~88%. Critically: above 31B parameters, hallucination rate dropped to zero on diagnostic tasks. The errors that remain look like the kind a careful clinician makes on a hard case — not the kind that destroys trust.
Source · Wang et al., UHN AI Hub, May 2026 · 188-model benchmark
HIPAA (US) · PDPA (Singapore + Thailand) · DPDP (India) · GDPR (EU). The same edge-only architecture clears all four.
Pre-loaded model + serving infrastructure + audit logging. Plug into clinic network. Done.
Any specialty-pharma platform, diagnostics distributor, or generics co selling into regulated markets fails or succeeds on the same data-residency + audit-trail question. On-device clinical AI sidesteps the entire category of failure modes.The clinic's compliance officer signs off in week one instead of month six. The hospital network's CTO stops blocking the AI rollout. The pharma partner stops asking "where does the patient data live?" because the answer is "the clinic, always."
Eli Lilly
Oral non-peptide GLP-1 — first small-molecule pill in the class
Eli Lilly
Triple agonist — GLP-1 + GIP + glucagon; class-breaking weight loss
Novo Nordisk
Amylin + GLP-1 combo — two complementary satiety signals
Eli Lilly (multiple programs)
Five-receptor agonist — next-gen after triple agonists
Amgen
Monthly-dosed GLP-1 antibody fusion — adherence game-changer
Biocon (India)
Generic tirzepatide — Indian patent challenge on evergreening grounds
Six monitors run continuously: FDA AdComm calendars, EMA CHMP meeting schedules, ClinicalTrials.gov Phase-3 update feeds, ADA / EASD / AHA / ASCO scientific-session abstracts, FDA Form-483 and Warning Letter feeds, and Indian CDSCO Subject Expert Committee (SEC) recommendation PDFs. When any monitor fires, the relevant patient cohort is auto-segmented for the clinic team.
Founder / CXO, 35–50
HNW urban, 35–55, F-skewed
Urban professional F, 40–55
35–55, M/F, doctor-flagged
NRI / expat, 30–50
Streams A–C earn at the platform layer (Omnis). Stream D earns at the operator layer (Omnis Digital Clinic). The four compound: more clinics → more protocols → more fulfilment → more multi-tenant data → better platform AI → higher LTV.
Compute and MD minutes scale with protocol volume. Watcher operations and customer success are fixed up to a clinic threshold. Click any line for the build-up.
The visual centerpiece. Read this like a spreadsheet, not a brochure. Left is Omnis Digital Clinic (operator-shape); right is Omnis (platform-shape). The two compound — see the bottom-line note.
Services-shape margins. Comparable to regional hospital-services operators at maturity.
SaaS-shape. Comparable to Veeva Vault unit economics at maturity, currency-adjusted for emerging-markets clinic pricing.
Two layers, same patient. Omnis Digital Clinic earns the services margin on the people the clinic serves directly; the Omnis platform earns the platform margin on every protocol routed through any partner clinic on the network. They compound — the more clinics on Omnis, the larger the multi-tenant dataset, the faster gate-pipeline AI improves, the higher the platform LTV.
Six comparables anchoring the multiple-band. Pure-DTC sets the ceiling for the operator layer; Veeva sets the ceiling for the platform layer. Indian listed CRDMOs + PE outcomes anchor the exit-path conversation.
$8–15B mkt cap
$2.35B FY25 rev · ~3–6× rev · DTC + Rx · US
$950M val · Aug 2025
$53M Series E · AI digital twin · metabolic
~$40B mkt cap · NYSE: VEEV
~12–15× revenue · ~35× EBITDA · B2B life-sci SaaS
$1.4B · Jan 2025 (Quadria)
~7× revenue · ~22–25× EBITDA · CRDMO
~$3.8B mkt cap · IPO Jul 2025
~18–20× revenue · listed Indian CRDMO
$3.01B EV · Jun 2025
5× MOIC · 36% IRR · Indian pharma PE exit
Public · NYSE HIMS · $2.35B FY25 (+59%)
Private · $7B (2022) · $1.03B raised
Private · biomarker DTC
Acquired · A$148M raised → H&H $1.5B (2024)
Six players. Premium tier maxed out. Building a 7th here is not a venture — it's a confirmation that we read the market wrong.
$649/yr · SG premium flagship
Launched Dec 2025 · Acquired Noviu
Private · Longevity screening + care
Prof. Andrea Maier-founded
Private · Preventive primary care + diagnostics
World's first public-hospital longevity clinic
Fragmented. Each player owns one slice — wearable, generic pharmacy, longevity clinic, coaching. None has built the consolidated stack.
$122M lifetime · Novo partnership Dec 2025
$53M Series E (Aug 2025) · $950M val
₹100cr debt + $120M raise · $550–600M val
$85M Series C (Aug 2025) · $410M val
$216M @ $710M (Apr 2024) · 90% haircut from $5.6B
Total Indian capital deployed in longevity 2016–2025: ~$50M across 12 rounds (Unisync Angels). Real category formation but pre-scale.
$3M seed Jun 2025 · $2.7M Feb 2026
$25M longevity fund (Oct 2025)
$31M Series B Jan 2026
Worth knowing the cluster exists, but not where Omnis plays. These operators sit in the legal-but-unenforced gap — Mumbai chains charge ₹20–40K/month for peptide stacks sourced through B2B "research peptide" wholesalers. The 24 March 2026 DCGI advisory (49-entity audit) signals the window closing. Omnis sells clean audit-grade substrate to whichever survivors operate post-enforcement; we don't build the consumer-facing peptide clinic ourselves.
Mumbai + Hyderabad
Gurugram / Pune / Delhi
Bangalore
Mumbai axis · Bandra / Khar / Andheri
Every player above owns one slice of the stack:
No one is the audit-grade clinical decision + monitoring + dispensing platform underneath all of them.
Omnis is that layer. The platform with on-device clinical AI, the gate pipeline, the watcher trio, and the provenance log — sold B2B to the clinics, telehealth platforms, and hospital networks that need to be HIPAA / PDPA / DPDP compliant by construction and to compete in the next 24 months as regulators tighten on the rest of the market.
The geographic question — which market we deploy first — is the optionality. The platform is geography-agnostic.
20+ approved ANDAs (Abbreviated New Drug Applications — the FDA filing every generic needs) · India + US + EU + RoW commercial presence · captive API + sterile manufacturing · the kind of platform institutional healthcare PE has been underwriting in the regulated-markets corridor
Omnis becomes the prescribing + monitoring layer for the specialty pipeline, plugged into clinic networks across regulated markets.
Diagnostic devices + lab equipment + IVD (In-Vitro Diagnostics — the lab tests done outside the body, like blood panels) distribution across SG / MY / India / Indonesia / Philippines / Thailand · built through serial acquisition · the recurring PE shape: distribution platform looking for a higher-margin software overlay
Omnis is the AI overlay on the distributor's lab volumes — every lab result fed into the gate pipeline before a clinician ever sees it.
Complex generics platform · US-India corridor · co-invested alongside global PE · products in complex injectables, hormones, oncology, ophthalmology
Omnis sits as the post-Rx monitoring layer — the platform ships, patients use, Omnis surveils.
Regional hub-and-spoke hospital chain · classic services-margin roll-up + operational improvement playbook · acquisition, professionalisation, surgical-mix expansion, then exit to pension or sovereign capital
The hospital-services roll-up is the pattern — Omnis is the platform that makes the next one faster to build.
Institutional healthcare PE underwrites companies that build durable specialty-pharma + healthcare-services infrastructure across regulated markets. Omnis is the platform layer that sits underneath every clinical decision those companies eventually need.
A regional diagnostics distributor's lab volumes need an AI overlay. A specialty-pharma platform's products need a prescribing + RWE layer. A complex-generics platform's output needs post-Rx surveillance. The next hospital-services chain needs operational substrate that scales clinical quality faster than human bandwidth allows.
Omnis is one stack that does all of those.
The open question is which of the four positions (or an adjacent one not named here) is the right first integration — and what the partnership shape would look like.
A specialty-pharma platform is the freshly-closed shape. A regional diagnostics distributor is the next-to-cycle shape. A complex-generics platform is the mid-cycle shape. A hospital-services roll-up is the rear-view shape but the pattern repeats. Of those four — or an adjacent position — where the first integration actually lives matters: customer (Omnis platform licensed into their clinic networks), partner (co-built next operating instance), or portfolio operator (Omnis Digital Clinic runs a flagship inside the broader healthcare-services thesis, with the platform earning underneath every clinic on the network).
Which one comes first changes what gets shipped in the next 6 months and where Omnis Digital Clinic operates first.
The institutional buyer profile already holds positions in SEA diagnostics distribution — devices, lab equipment, IVD across the SG / MY / India / Indonesia / Philippines / Thailand corridor. Every lab result delivered through that footprint is a data point Omnis could interpret + protocol-recommend on. That's the most natural pilot partnership for the platform layer specifically.
A 60-minute conversation with the operating team of a comparable distributor would tell us whether the integration mechanic mapped here matches how distribution actually delivers value to clinic customers — or whether the real shape is something else.
We came in thinking consumer longevity. After watching specialty-pharma and healthcare-services capital close into the regulated-markets corridor, we think the right shape is something else. We're here to find it.
What we're certain on: the four-piece platform — gate pipeline, watcher trio, on-device clinical AI, frontier-compound readiness. The platform exists, it survives scrutiny, and it slots into the regulated-markets infrastructure thesis institutional healthcare capital is already underwriting.
What we're open on: geography, operating-co posture, business-model emphasis, partnership structure. Those answers come out of the right partnership conversations.