WHY THIS, WHY US, WHY NOW

A federal deadline is closing
the API half of prior auth.
We built for the half it leaves behind.

CMS-0057-F requires impacted payers to run a FHIR Prior Authorization API by 1 January 2027. That rule is the single most important fact about this market — and reading it carefully is the whole thesis. It does not cover every payer, it does not cover every request, and it does not remove the human review at the end. It compresses the clean, structured half of prior authorization into APIs, and it leaves a smaller, denser, permanently manual half behind. DeskGate is built for that half, and the transition window is now.

The thesis in five sentences

The regulated segment gets APIs by 2027; ERISA self-funded employer plans, drug prior authorization, and commercial non-exchange plans do not. During the transition, coordinators work a mixed environment — some payers structured, most not — which is the most expensive configuration a desk can be in. We already have a working in-session runtime, not a specification. We reached healthcare by building one engine and testing it across six verticals until the evidence pointed here. What we are raising against is the cost of putting that runtime on one or two real desks and publishing what it measures.

What the rule actually leaves on the desk
Segment or workflow Covered by CMS-0057-F? Residual portal work after 2027
Medicare Advantage, Medicaid / CHIP, QHPs on federal exchanges Yes — FHIR PA API required Shrinking. Structured exchange replaces the portal for covered requests.
ERISA self-funded employer plans Exempt from the rule Persists indefinitely. Portal and fax remain the mechanism.
Drug prior authorization Excluded from the PA API provisions Persists. Specialty and step-therapy work stays manual.
Requests routed to clinical review API carries the request, not the decision Persists. Documentation, attachments, and appeals still assembled by hand.
Denials, appeals, status chase Partially — decision reasons required, workflow not Persists. The rebuttal packet is still human work.
2026–2028 transition period Phased, payer-by-payer, frequently late Peak pain. Mixed environments cost more than uniformly manual ones.
Why this is a window and not a permanent market
We are explicit about this because it determines the shape of the company. The regulated slice of prior authorization is being automated by mandate, and we are not trying to compete with a federal API. Our addressable work is the exempt and excluded remainder plus the multi-year transition — which is large, urgent, and getting more concentrated rather than less. A desk running a mixed book in 2027 has fewer manual payers but no less confusion about which is which. That is a better market to serve than today's uniform mess, not a worse one.
🔑 Five things that are true of us and rarely true at this stage
01 · BUILT, NOT SPECIFIED

The runtime exists and runs

embed.js, the operator overlay, screen-grounded staging, the evidence hasher, and the dossier forge are working software, not a roadmap. A pre-seed cheque here funds distribution and proof, not discovery of whether the thing can be made. That is a materially different risk profile from the median cheque at this stage.

02 · CONVERGED ON EVIDENCE

The wedge is a conclusion, not a hunch

One engine was built and pointed at six verticals — real estate valuation, critical minerals, model-economy intelligence, prospect research, general portal work, and healthcare administration. Healthcare prior-auth won on observed friction, buyer urgency, and willingness to engage. We are not guessing where we fit; we tested and discarded the alternatives at our own cost.

03 · APPROVABLE BY DESIGN

The human gate is strategy, not a limitation

Competitors chasing full autonomy meet a compliance committee that will not accept unattended, un-audited filing against a payer. We stop before the irreversible click on purpose. That constraint is what lets a desk say yes without a risk review that kills the pilot. It costs us the "fully autonomous" headline and buys us the deployment.

04 · INSTRUMENTED DEMAND

We can find our own buyers

Our targeting engine ranks live operators with residual portal work — scored by payer surface, workflow type, and desk shape, with noise filtered out. Most companies at this stage cannot name their first twenty customers. We can generate that list on demand and show you the scoring logic behind it.

05 · WE SELL A MEASUREMENT, NOT A PROMISE

The commercial structure removes the buyer's risk before it removes ours

We withdrew the pass/fail savings SLA deliberately: it made us the underwriter of the client's cohort, portal family, and exception rate — variables we do not control. What replaces it is a 14-day frozen-cohort ledger on one payer family, with mapping fees credited against Year 1 on conversion. A desk can say yes without betting on a vendor's projection, and an investor gets an audited operating number instead of a testimonial. This is why the first desks are reachable at all.

📈 What compounds after the first desks
PER DESK

Portal behaviour corpus

Each measured lane teaches us how a specific payer family behaves: field semantics, what attachments get accepted, which rejections are formatting versus substance, where staging breaks. This only comes from live desks. A competitor with identical architecture starts this at zero.

PER CASE

The evidence ledger

Hashed provenance of what was staged, by whom, and what the payer returned. As a desk accumulates months of audit-defensible record, that history becomes something they depend on and would have to rebuild elsewhere — a switching cost that grows without us building anything new.

PER RELATIONSHIP

Reference desks

In healthcare administration a named peer describing measured results outsells any feature list. The first one or two are the hardest thing to obtain and the most valuable thing to hold. This is precisely where an investor with a healthcare portfolio changes our trajectory rather than merely funding it.

On the architecture itself
In-session execution, screen grounding, human gating, and local-only PHI handling are the correct engineering answers to hospital VPNs, MFA, bot management, and compliance review — and other serious teams are reaching the same conclusions. We do not present those four choices as a barrier to entry. They are why we work today where cloud RPA does not. The durable position is built on top of them, from the three assets above, and it is earned desk by desk rather than claimed in advance.
⚠️ The case against us, stated by us

We would rather you test these than discover them:

  • If FHIR adoption runs faster and broader than the rule requires, our addressable residue shrinks sooner. We think the exempt segments make that a long tail rather than a cliff, but it is the central risk and we do not dismiss it.
  • The architecture is replicable. A funded team could build these four pillars in a quarter. Our answer is speed to live desks and what accumulates there — not a technical barrier.
  • We have no paying customers yet. No logos, no revenue, no completed measurement. That is precisely what this round is for, and we would rather say so than dress up a pilot as traction.
  • Healthcare sales cycles are slow and reference-gated, which is why our commercial structure leads with a 14-day measurement and why a portfolio introduction is worth more to us than the cheque alone.
  • We are a small team. Two founders. The runtime is built; the distribution is not.
What we are not claiming
  • — That we replace Epic, a clearinghouse, or any system of record.
  • — That we make clinical or medical-necessity determinations. The coordinator does.
  • — A guaranteed savings figure. We publish a measured ledger instead of an SLA.
  • — That our architecture cannot be copied. It can, and probably will be.
  • — That any of this is proven at scale. It is not, yet. That is the point of the first desks.

Sources for the claims on this page

  1. CMS-0057-F scope and timing: impacted payers — Medicare Advantage organisations, state Medicaid and CHIP fee-for-service, Medicaid and CHIP managed care, and QHP issuers on federally facilitated exchanges — must implement the Prior Authorization, Patient Access, Provider Access, and Payer-to-Payer FHIR APIs generally by 1 January 2027, with operational decision-timeframe requirements already in force since 1 January 2026. (Ref: CMS — Interoperability and Prior Authorization Final Rule) [1]
  2. Exemptions and exclusions: the rule applies to government-regulated payers; self-funded employer plans under ERISA are not impacted payers, and the Prior Authorization API provisions cover medical items and services rather than drug prior authorization. The rule also does not mandate real-time approval — requests may still route to clinical review. (Ref: CMS-0057-F final rule text) [2]
  3. Why datacenter automation struggles on authenticated payer surfaces: commercial bot-management systems use TLS fingerprinting, IP reputation, and behavioural heuristics that flag automated headless browsers. (Ref: Cloudflare bot mitigation overview) [3]
  4. Why screen grounding outperforms static DOM paths: web-agent benchmarks such as Mind2Web and WebVoyager report that multimodal grounding sustains execution across interface changes where raw XPath selectors degrade sharply. (Ref: WebVoyager: Building an End-to-End Visual Web Agent) [4]
  5. Data-minimisation posture: under 45 CFR § 164.312 technical safeguards, in-memory processing that avoids persistent cloud storage of ePHI removes a class of multi-tenant exposure. REA v1.2 generates a SHA-256 hash over the portal response payload, timestamp, and operator confirmation. (Ref: HHS HIPAA security standards) [5]

References describe the regulatory and technical environment. They are not evidence of DeskGate's own performance — that is what the 14-day measurement exists to produce, and we have not produced it yet.

The ask

A pre-seed SAFE and an introduction to one or two authorisation desks you already know.

The capital funds the deployment. The introduction is what converts a working runtime into a published ledger — and that ledger is what prices the next round. If you hold a healthcare portfolio, you can do something for this company that money alone cannot.

Open the conversation