The investor thesis

Hospitals remember patients.They forget decisions.

Health systems keep systems of record for care, claims, finance, staffing, and transactions. None of them holds what leadership knew, what it believed, what it decided and why, who owned the response, or what happened afterward.

sanalytics is the AI operating executive for health systems: an always-on executive teammate that connects evidence, judgment, decisions, coordinated action, outcomes, and institutional memory.

Built end to end. Preparing the first health-system deployment.Synthetic hospital data

Product
End-to-end controlled workflow builtTested on synthetic hospital data
Executive discovery
50+Hospital and health-system executive conversations
Core U.S. opportunity
~$2.25BBottom-up modeled opportunity, not current revenue
Next proof
First health-system deploymentNot yet deployed in a live hospital

The non-obvious insight

Data-rich.Intelligence-fragmented.Decision-memory-poor.

The scarce thing inside a health system is not another metric. It is the connective record of how leadership reasoned, what it chose, and what the choice produced.

Systems of record exist for

  • Patients
  • Claims
  • Finance
  • Staffing
  • Transactions

What no system persistently keeps

  • What leadership knew
  • What leadership believed
  • What it decided, and why
  • Who acted
  • What happened next

Hospitals have systems of record for almost everything except management judgment.

That is the category sanalytics is building for: a system of record for management judgment. It describes where the product sits in the stack, not a regulatory designation.

Why the entry wedge works

Start where the fragmentation is already measurable.

Nobody buys an AI executive on day one. They buy help with the meeting they already run every month, which happens to exercise the entire loop the company is built around.

The entry point

The recurring executive operating review

One contained workflow that carries evidence, judgment, decision, ownership, and measurement in a single cycle.

The questions it starts with

  • CMI and DRG mix
  • LOS vs expected GLOS
  • Facility and provider variance
  • Discharge pathways
  • Throughput and capacity
  • Documentation and coding
  • The financial movement underneath
  • Recurring

    It runs every month, so the product earns a standing seat instead of a one-time look.

  • Cross-functional

    It pulls six executive seats into the same set of numbers at the same time.

  • Evidence-rich

    The underlying records already exist in structured hospital systems.

  • Financially legible

    Movement in the review shows up in operating performance.

  • Measurable

    The decision has a baseline, an owner, and a next review date.

  • Already executive work

    Nothing new has to be adopted for the workflow to exist.

The operating review is the entry point, not the ceiling.

The product loop

One operating problem, carried end to end.

The controlled product already moves a single hospital operating issue through five stages without losing the evidence, the reasoning, or the person accountable for the response.

  1. Ask

    An operating question answered with governed evidence.

    Leaves behindA scoped question with its definition, window, and known gaps.

  2. Brief

    The same issue read from the perspectives of multiple executive seats.

    • CEO
    • CFO
    • COO
    • CMO
    • CNO
    • CIO / CMIO

    Leaves behindSix readings drawn from one evidence base.

  3. Decide

    A bounded, pressure-testable response with explicit assumptions and human authority.

    Leaves behindOptions carrying what they assume and what would break them.

  4. Act

    One approved decision translated into role-specific draft work, with evidence attached and nothing sent autonomously.

    Leaves behindDraft work packages held at an explicit approval gate.

  5. Learn

    Baseline, decision, owner, success measure, and the subsequent outcome stay on the record.

    Leaves behindA record the next review opens with.

sanalytics · Operating memoryIllustrative product experience · Synthetic data

The trail one decision leaves

Adult-medicine length of stay running above expected.

Nothing here is a dashboard state. It is the reasoning, the authority, and the result held together as one record.

Baseline agreed
5.8 days
Decision
Coordinate the discharge pathway review
Accountable owner
Operations leadership
Success measure
Stay gap, 14-day review
Outcome status
Awaiting post-intervention observation

Observed association · Not a confirmed cause · Human approval required before anything moves

Every operating decision leaves a measurable trail.

What is actually built

A governed system, not a model pointed at hospital data.

The reason this is defensible technically is the same reason it is sellable in healthcare: the reasoning is constrained, the calculation is deterministic, and the authority stays human.

  • Governed hospital-style data intake and validation
  • Deterministic operating metrics and lineage
  • Evidence-bound investigation
  • Executive C-suite briefing
  • Bounded decision pressure-testing
  • Draft decision and action workflow
  • Structured operating memory
  • Human-review and refusal controls

Built and tested end to end in a controlled synthetic environment. Live hospital deployment is the next milestone.

  1. Modelspropose.
  2. Contractsconstrain.
  3. Toolscalculate.
  4. Evidenceproves.
  5. Humansdecide.
  1. Structured hospital data is validated before anything is computed.

  2. Governed metrics are calculated deterministically, not generated.

  3. Model reasoning stays bound to the evidence actually available.

  4. Conclusions the evidence cannot support are refused.

  5. Decisions and coordinated action stay under human control.

  6. Baselines, measures, decisions, and outcomes stay attached to memory.

Those design-stage controls are real and running today. Production controls are a separate body of work: enterprise role-based access, retention and deletion policy, production-scale audit logging, business-associate agreements and customer security review, monitoring, and deployment-specific hosting are being scoped as deployment requirements. They are not completed certifications, and sanalytics does not claim any.

The compounding thesis

EHRs remember care.BI remembers metrics.sanalytics remembers management judgment.

Defensibility does not come from the model. It comes from what the product is left holding after each operating cycle closes.

  1. Question
  2. Evidence
  3. Decision
  4. Action
  5. Outcome
  6. Memory

Each closed loop returns as context for the next question.

As deployment history accumulates, sanalytics can gain institution-specific operating context: how this particular health system frames a problem, which evidence it treats as decisive, who it holds accountable, and what actually moved afterward. The next operating question inherits the reasoning and the outcomes of the ones before it.

This is a compounding thesis, not a claim of scale today. sanalytics does not hold years of deployed institutional decision history. It holds the architecture that would accumulate it, and one controlled environment where the loop already closes end to end.

  • Defines problems
  • Weighs evidence
  • Makes decisions
  • Assigns responsibility
  • Responds to intervention
  • Measures success

Why now

AI can finally hold the context.Hospitals finally need it governed.

  1. One model can hold the whole operating picture

    Modern models reason across language, structured data, documents, and organizational workflows at the same time. The executive layer was previously the part software could not read.

  2. Hospital pressure keeps rising

    Margins, capacity, workforce constraints, and reimbursement complexity have made executive coordination harder at exactly the moment there is more information to hold.

  3. Trust is the gating layer

    Healthcare cannot safely buy opaque autonomous reasoning. Evidence, provenance, defined authority, human review, and bounded execution decide what gets deployed.

The winner will combine AI-level breadth with hospital-grade evidence and control.

Market and business model

Built bottom up from hospitals that already run the review.

The model starts from published hospital counts and a modeled mature annual value per hospital-equivalent. It is an opportunity construction, not a pipeline and not a forecast.

5,121U.S. community hospitals

  1. System-affiliated

    3,567hospitals

    ~$500Kmodeled mature annual value

    ~$1.78B

  2. Independent

    1,554hospitals

    ~$300Kmodeled mature annual value

    ~$466M

Core U.S. annual opportunity

≈ $2.25BBottom-up modeled opportunity, not current revenue.Modeled mature ACV assumptions. Not current contracts.

Why it gets bigger

The number above only models the initial U.S. community-hospital market using the current mature-ACV assumptions. It excludes the broader executive-layer vision.

  • Large systems buy at the enterprise level rather than per site.
  • Value scales with the number of facilities in scope.
  • Value scales with the data domains the product is trusted with.
  • Value scales with the executive workflows it carries.
  • Multi-hospital scope becomes a materially larger account.
  • Future operating domains extend well past the initial wedge.

Source: American Hospital Association, Fast Facts on U.S. Hospitals. Counts rounded as published.

Category positioning

Strong companies validate the pieces.sanalytics connects the executive loop.

  • Hospital operations AI

    Qventus · LeanTaaS

    These companies validate that AI can improve major hospital operating workflows such as discharge, operating-room scheduling, and capacity. sanalytics is built around a different unit of work: the cross-functional executive loop, and what happens after the operating review.

  • Systems of record

    Epic · Oracle Health

    They hold clinical and transactional truth, and they prove that hospitals buy platforms that remember. What they remember is care and transactions. Management judgment, ownership, and the trail of executive decisions are not the record they primarily maintain.

  • The practical incumbent

    BI teams · internal analytics · consultants

    This is how the executive operating review usually runs today: dashboards, ad hoc analysis, recurring decks, and manual follow-through. The work is real and often excellent. The reasoning trail is what fragments after the meeting.

Where sanalytics sits

  • Cross-functional executive context
  • Governed evidence
  • Decision workflow
  • Coordinated follow-through
  • Measurement
  • Operating memory

The claim is not that any of these categories is deficient. It is that evidence, decisions, follow-through, and memory currently live in different places, and nothing holds them together.

Founder

I didn’t start with AI and look for a healthcare problem.I started inside hospitals.

Luke Green has been inside hospital operations since high school, moving from floor-level workflow to quantitative hospital operations, then to the executive conversations that shaped what sanalytics became.

  1. Frontline

    Began working in emergency departments while still in high school.

  2. Responsibility

    Became a lead medical scribe and trainer.

  3. Operations

    Gained exposure to hospital administration, workflow, and billing.

  4. Quantitative work

    Led research and analytics for an intensivist / Tele-ICU operation serving hospitals across Greater Los Angeles, including hospital outcomes work.

  5. Executive discovery

    50+ conversations across hospital and health-system leadership.

  6. Execution

    Built sanalytics and its current operating architecture end to end as a solo founder.

I didn’t start with AI and search for a problem. I kept encountering the same operating gap inside hospitals, and sanalytics is what I built in response.

At UCSB, Luke completed the pre-med curriculum while studying biology, economics/accounting, and philosophy.

Customer learning became product

The executives changed what got built.

  1. Early

    Broad hospital analytics and operating intelligence.

  2. What executives kept asking about

    • Inpatient variance
    • OR utilization
    • Imaging capacity
    • Throughput
    • Financial performance
    • How decisions actually moved after the review
  3. The realization

    The deeper problem was not another metric. It was preserving the path from evidence to judgment to decision to ownership to measurement.

  4. The result

    sanalytics became the evidence architecture, the closed-loop decision workflow, the operating memory, and the AI executive experience.

The wedge becomes the platform

Today, sanalytics helps leadership understand an operating problem.Eventually, it understands the health system itself.

  1. Today

    One operating problem. sanalytics investigates it, briefs leadership, structures the decision, coordinates approved follow-through, and preserves the trail.

    Running end to end in a controlled synthetic environment.

  2. Next

    More operating domains: finance, capacity, workforce, clinical performance, strategy, and market intelligence.

    Planned expansion, not current capability.

  3. Eventually

    Persistent context across the whole health system, so leadership can ask the questions nobody currently has a system to answer.

    The long-term product ambition.

Illustrative future questions

  • What unexplained problem should we focus on today?
  • Which intervention worked last time?
  • Where is a competitor becoming dangerous?

None of these are live capabilities. They describe what a persistent executive context layer would make askable.

The next proof point

If this thesis is right, the executive layer becomes a new system of memory for the health system.

The first deployment is deliberately narrow. One health system, one recurring review, one measured loop, run end to end and then repeated.

  1. Scoped data
  2. Recurring executive workflow
  3. Named executive sponsor
  4. Evidence and decision loop
  5. Agreed success measure
  6. Repeatable readout

Prove one loop end to end inside one health system, then repeat it.

The controlled product is built. The next proof point is the first health-system deployment.

Please do not send PHI, patient-level data, credentials, or confidential hospital information by email.