Travi is architected as a Clinical Decision Support (CDS) tool, not a diagnostic device — a distinction the FDA has explicitly codified in its 2019 CDS guidance and the 21st Century Cures Act. CDS tools that present information for a clinician to independently review and act upon are exempt from FDA device regulation. Travi does not diagnose, treat, or prescribe — it extracts, organizes, and surfaces the patient’s own discharge plan and flags non-adherence for a clinician to evaluate.
\nThe specific design choices that maintain this positioning: (1) all clinical outputs are presented as information, not recommendations; (2) the clinician brief explicitly labels every item as “for your review”; (3) no Travi output triggers an automated clinical action without human review; (4) the escalation system alerts a care coordinator, who then decides whether to intervene.
\nWe have engaged healthcare regulatory counsel to review the architecture against FDA’s Digital Health Center of Excellence guidance and have scoped a 510(k) pathway as a contingency. Our current assessment is that FDA clearance is not required for the initial product, but we are monitoring the regulatory environment actively.
\n" }, { "cat": "regulatory", "q": "How does Travi handle HIPAA? Who signs the BAA?", "search": "How does Travi handle HIPAA Who signs the BAA Business Associate Agreement PHI data privacy security", "a": "Travi operates as a Business Associate under HIPAA and executes a Business Associate Agreement (BAA) with every health system customer before any PHI is accessed. The BAA defines Travi’s permitted uses of PHI, breach notification obligations, and data destruction requirements at contract termination.
\nInfrastructure is hosted on AWS HIPAA-eligible services (Amazon RDS, S3, EC2, Lambda) under Travi’s own BAA with AWS. PHI is encrypted at rest (AES-256) and in transit (TLS 1.3). Access controls follow the principle of least privilege with role-based access and full audit logging.
\nTravi is pursuing SOC 2 Type I certification at month 6 and SOC 2 Type II at month 14. These certifications are required by most health system procurement teams and are on the critical path to closing the first paid contract.
\n" }, { "cat": "regulatory", "q": "What happens if CMS changes the HRRP penalty structure?", "search": "What happens if CMS changes HRRP Hospital Readmissions Reduction Program penalty structure reimbursement TEAM model", "a": "HRRP has survived every legislative cycle since its creation in 2012 and has bipartisan support as a cost-reduction mechanism. The probability of elimination is low. However, even if HRRP were significantly reduced, Travi has three independent financial drivers that do not depend on HRRP:
\nThe HRRP penalty is the primary sales conversation opener, but it is not the only ROI driver. A hospital that has deployed Travi and seen readmission rates decline has strong operational and clinical reasons to continue regardless of the penalty structure.
\n" }, { "cat": "regulatory", "q": "How does Travi manage LLM hallucination risk in a clinical context?", "search": "How does Travi manage LLM hallucination risk clinical context AI safety 7-gate validation pipeline", "a": "This is the most important technical risk question, and we have designed the entire plan compiler architecture around it. The answer is the 7-gate validation pipeline — a deterministic post-processing layer that every LLM output must pass before it enters the clinical workflow:
\nAny output that fails gates 4–7 is held for human review before entering the clinical workflow. The pipeline is designed so that a hallucination that passes all 7 gates would need to be both clinically plausible and consistent with the source document — a significantly higher bar than raw LLM output.
\n" }, { "cat": "timing", "q": "Why is now the right time to build this? Why not 3 years ago or 3 years from now?", "search": "Why is now the right time to build Travi why not 3 years ago FHIR LLM TEAM model window", "a": "Three conditions converged in 2023–2026 that make this the precise window:
\n| Condition | Before 2023 | Now |
|---|---|---|
| LLM capability for clinical text | Insufficient reliability; hallucination rates too high for clinical use | GPT-4 class models achieve reliable structured extraction with a validation pipeline |
| FHIR API access | EHR vendors not required to expose FHIR APIs; integration required custom HL7 work costing $500K+ | CMS mandate requires all EHRs to expose FHIR R4 APIs; integration tractable at startup scale |
| Financial urgency (HRRP + TEAM) | HRRP penalties existed but were capped at 1–2%; TEAM model did not exist | HRRP penalties up to 3% of all Medicare FFS; TEAM model launched January 2026 at ~750 mandatory hospitals |
Three years ago, the technical problem was unsolvable at startup scale. Three years from now, a well-funded competitor will have the outcome data moat that Travi is building today. The window is now.
\n" }, { "cat": "timing", "q": "How large is the addressable market, and how does Travi get to $100M ARR?", "search": "How large is the addressable market TAM path to 100 million ARR revenue hospitals", "a": "The immediate addressable market is the ~3,000 US hospitals subject to HRRP penalties, with a particular focus on the ~750 hospitals in the mandatory TEAM model. At a blended ACV of ~$394K per hospital (platform fee + per-episode revenue at 2,500 episodes/year), the TAM for this segment alone is approximately $1.2B ARR.
\nThe path to $100M ARR requires approximately 250 hospital customers at the blended ACV — roughly 8% penetration of the HRRP-penalized hospital market. At the base case growth trajectory (60 customers by Year 5), Travi reaches ~$43M ARR in Year 5 and $100M ARR in Year 7–8.
\nThe expansion opportunity beyond hospitals includes physician group practices (TCM billing), Medicare Advantage plans (readmission risk reduction), and post-acute care facilities (SNF-to-home transitions) — each of which represents an additional $500M+ TAM layer that Travi can address with the same core platform.
\n" }, { "cat": "timing", "q": "The digital health market has been difficult since 2022. Why will Travi be different?", "search": "digital health market difficult 2022 downturn correction why will Travi be different B2B clinical AI", "a": "The digital health correction of 2022–2024 hit companies with three specific characteristics: consumer-facing models with high CAC, reimbursement models dependent on telehealth waivers, and growth-at-all-costs unit economics. Travi has none of these characteristics.
\nTravi sells to hospital CFOs and CMOs on a quantifiable ROI case (HRRP penalty avoidance). The sales motion is B2B enterprise, not consumer. The revenue model is SaaS with per-episode fees — not reimbursement-dependent. And the unit economics are designed for gross margin expansion as the outcome dataset grows, not deterioration as CAC compounds.
\nThe companies that survived and thrived through the correction — Abridge, Nabla, Suki — are all B2B clinical AI companies with clear workflow value and enterprise sales motions. Travi is in that category, not the consumer wellness or telehealth categories that were most affected.
\n" }, { "cat": "pilot", "q": "What does the design partner pilot look like? What are the terms?", "search": "What does the design partner pilot look like terms fee structure 6 months FHIR integration", "a": "The design partner pilot is a 6-month paid engagement structured in two phases:
\n| Phase | Duration | Scope | Fee |
|---|---|---|---|
| Phase 1 — Integration & Onboarding | Months 1–2 | FHIR integration, workflow mapping, care coordinator training, first 10–20 patients enrolled | $15,000 (implementation fee) |
| Phase 2 — Live Pilot | Months 3–6 | 50–150 patients enrolled across 1–2 DRG groups (HF, pneumonia, or COPD), full outcome tracking, weekly check-ins | $25,000–$40,000 (pilot fee) |
At the end of the 6-month pilot, the health system receives a 30-day readmission outcome report comparing the Travi cohort to a matched historical control. If the outcome data is positive, the pilot converts to a full annual contract at the standard ACV. Design partner selection criteria: HRRP-penalized hospital with documented readmission rate above national average; Epic or Oracle Health EHR; identified physician and care coordinator champions; willingness to share de-identified outcome data for research publication.
\n" }, { "cat": "pilot", "q": "What is the success metric for the pilot? What counts as a win?", "search": "What is the success metric for the pilot readmission reduction 30-day outcome win criteria", "a": "The primary success metric is 30-day all-cause readmission rate in the Travi cohort vs. a matched historical control (same DRG, same time of year, same risk stratification). Travi targets a minimum 15% relative reduction as the threshold for a compelling outcome report.
\nSecondary metrics tracked during the pilot:
\nA pilot is considered a commercial success if the primary metric is met and the health system signs a renewal contract. A pilot is considered a research success if the outcome data is sufficient to support a peer-reviewed publication, which Travi will pursue regardless of the commercial outcome.
\n" }, { "cat": "pilot", "q": "What if the pilot does not show a statistically significant readmission reduction?", "search": "What if the pilot does not show statistically significant readmission reduction failure scenario", "a": "This is the right question to ask, and the honest answer is: a single 6-month pilot with 50–150 patients will not be statistically powered to detect a 15% relative reduction at 95% confidence. The pilot is designed to generate a directional signal and a compelling case study, not a randomized controlled trial.
\nThe clinical evidence base for care transitions interventions is already strong — the BOOST, RED, and BRIDGE trials have demonstrated 20–30% readmission reductions with structured post-discharge support. Travi is applying AI to make those interventions more scalable and consistent. The pilot is validating the implementation, not the underlying clinical hypothesis.
\nIf the pilot shows no directional improvement, the most likely explanations are implementation fidelity (care coordinator adoption, patient engagement rates) rather than a failure of the underlying clinical model. A failed pilot is a learning event, not a company-ending event — it informs the product roadmap and the next pilot design.
\n" }, { "cat": "pilot", "q": "How long does it take to get a health system from LOI to live pilot?", "search": "How long does it take to get a health system from LOI to live pilot procurement timeline legal", "a": "The typical timeline from LOI to first patient enrolled is 4–6 months:
\n| Stage | Duration | Key Activities |
|---|---|---|
| Legal & procurement | 4–8 weeks | BAA execution, MSA negotiation, IT security review, procurement approval |
| FHIR integration | 4–6 weeks | EHR API credentialing, FHIR endpoint mapping, data validation testing |
| Workflow configuration | 2–3 weeks | DRG selection, escalation rule configuration, care coordinator dashboard setup |
| Training & go-live | 1–2 weeks | Care coordinator training, physician champion briefing, first patient enrollment |
Travi uses a two-component SaaS pricing model:
\n| Component | Amount | What It Covers |
|---|---|---|
| Annual platform fee | $120,000–$180,000/year | Unlimited users, EHR integration, dashboard, clinician brief, support |
| Per-episode fee | $90–$110 per enrolled episode | AI plan extraction, patient task execution, escalation monitoring for one 30-day episode |
At a typical hospital with 2,500 episodes per year, the blended ACV is approximately $370,000–$455,000. The per-episode fee aligns Travi’s incentives with the hospital’s outcome goals. The pricing is benchmarked against the financial value delivered: at 2,500 episodes per year with a 15% readmission rate and a 20% relative reduction, Travi prevents approximately 75 readmissions per year — a $1.1M–$1.9M annual value against a $370K–$455K annual cost.
\n" }, { "cat": "business", "q": "What is the sales cycle, and who is the economic buyer?", "search": "What is the sales cycle who is the economic buyer CFO CMO enterprise B2B hospital", "a": "The economic buyer is the hospital CFO, with the CMO or VP Quality as the clinical champion. The CFO controls the HRRP penalty budget and has a direct financial incentive to reduce readmissions. The CMO provides clinical credibility and physician champion access. Both need to be engaged for a deal to close.
\nThe typical sales cycle is 9–18 months from first meeting to signed contract, depending on the health system’s procurement complexity. Large IDNs (10+ hospitals) have longer cycles (12–18 months) but larger contract values. Community hospitals (1–3 hospitals) have shorter cycles (6–9 months) and are the primary target for the first 3–5 design partners.
\nThe sales motion is relationship-driven, not inbound. The founding team’s existing relationships with hospital leadership and clinical advisory board connections are the primary pipeline source at the seed stage. A VP of Sales is planned for hire at month 4 to build a systematic outbound pipeline.
\n" }, { "cat": "business", "q": "What is the gross margin, and how does it improve over time?", "search": "What is the gross margin how does it improve over time unit economics LLM cost integration amortization", "a": "Gross margin at the design partner stage is approximately 58–62%, reflecting the high cost of early integration work and customer success support. As the customer base scales, gross margin expands toward 70–76% through three mechanisms:
\n1. Business model misalignment. Epic’s revenue model is based on inpatient workflow licensing. The discharge event is the end of Epic’s value chain. Building a post-discharge product would require Epic to take on clinical liability for outcomes that occur outside the hospital — a risk profile inconsistent with their business model and their contracts with health systems.
\n2. Development cycle mismatch. Epic operates on 18–24 month release cycles with a highly conservative change management process. By the time Epic could design, build, test, and release a competing feature, Travi will have 2–3 years of outcome data, published clinical evidence, and deeply embedded relationships at 20–30 health systems. The switching cost at that point is high.
\n3. Organizational incentives. Epic’s engineering resources are allocated by customer demand. Health system CIOs — Epic’s primary buyers — are not asking Epic to build a post-discharge AI agent because that is not a CIO problem. It is a CFO and CMO problem. Travi sells to a different buyer within the same institution.
\n" }, { "cat": "competition", "q": "What stops a well-funded competitor from copying Travi’s product?", "search": "What stops a well-funded competitor from copying Travi product moat data defensibility IP patent", "a": "A competitor can copy the product architecture from a description. They cannot copy the outcome dataset. Every episode Travi manages generates a labeled outcome record — discharge plan, patient actions, 30-day result — that trains and validates the predictive models that make Travi’s escalation triggers more accurate over time. This dataset is the moat, and it grows with every patient.
\nA competitor entering the market in Year 3 faces a Travi that has processed 50,000+ episodes, has 3 years of outcome data, has published clinical evidence, and has deeply embedded relationships at 20+ health systems. The cost of replicating that is not engineering cost — it is time. And time is the one thing a well-funded competitor cannot buy.
\nAdditionally, Travi is filing provisional patents on the 7-gate validation pipeline and the execution graph architecture. These are not the primary moat — the data is — but they add a layer of formal IP protection that increases the cost of a direct copy.
\n" }, { "cat": "team", "q": "Why is this team the right team to build Travi?", "search": "Why is this team the right team to build Travi founder market fit clinical access technical depth", "a": "The founding team has three specific advantages that are difficult to replicate:
\nThe biggest execution risk is health system procurement speed. Hospital procurement cycles are notoriously slow — legal review, IT security assessment, clinical governance approval, and budget cycle alignment can each add weeks or months to a deal that the clinical champion is ready to sign. This is the primary reason the seed round runway is 18–20 months rather than 12.
\nWe are managing this risk in three ways: (1) targeting community hospitals and regional health systems for the first 2–3 design partners, where procurement cycles are 6–9 months rather than 12–18; (2) engaging healthcare legal counsel to pre-negotiate standard BAA and MSA templates that reduce the legal review cycle; (3) pursuing retrospective data agreements in parallel with prospective pilot negotiations, so that outcome data collection begins before the full procurement cycle completes.
\nThe second execution risk is physician engagement. We are addressing this by designing the clinician brief for zero-friction consumption — no login, no new workflow, delivered through the EHR’s existing messaging system — and by requiring a physician champion as a design partner selection criterion.
\n" }, { "cat": "team", "q": "What does success look like at 18 months (Series A raise)?", "search": "What does success look like at 18 months Series A raise milestones ARR customers SOC 2", "a": "At 18 months, a Series A-ready Travi has:
\nTravi is raising a $4–5M seed round. The round is structured as a SAFE (Simple Agreement for Future Equity) with a valuation cap of $[X]M and a [X]% discount. The seed round provides 18–20 months of runway at the projected burn rate, with a target Series A raise at month 18 upon achieving the milestones described above.
\nThe use of funds is detailed in the Use of Funds page — the primary allocation is engineering (42%), followed by clinical operations (20%), commercial (15%), compliance (13%), and infrastructure (10%). We are targeting a small number of lead investors with healthcare IT expertise and a track record of supporting B2B clinical AI companies through the design partner phase. Strategic angels with health system relationships are particularly valuable at this stage.
\n" }, { "cat": "financials", "q": "What is the pre-money valuation, and how did you arrive at it?", "search": "What is the pre-money valuation how did you arrive at it comparable seed rounds clinical AI", "a": "The pre-money valuation is $[X]M, based on three reference points:
\nThe primary exit path is a strategic acquisition in the $200M–$800M range within a 7–10 year horizon. The most likely acquirers are Epic, Oracle Health, Optum/UnitedHealth, CVS/Aetna, and large integrated delivery networks — all of whom have a clear strategic rationale for acquiring Travi once the outcome data moat is established. The Exit Strategy page documents the full acquirer landscape and comparable transactions.
\nAn IPO is a secondary path if Travi reaches $50M+ ARR with strong growth and the public market window is favorable. Seed investors should expect a 7–10 year hold period with a target return of 20–40x on a strategic exit.
\n" }, { "cat": "financials", "q": "What is the single biggest risk to the company, and what keeps you up at night?", "search": "What is the single biggest risk to the company what keeps you up at night honest answer", "a": "The honest answer: procurement speed combined with a long sales cycle is the risk that could cause the company to run out of runway before achieving the Series A milestones. Health system procurement is slow and unpredictable. A deal that the clinical champion is ready to sign in month 6 can sit in legal review for 4 months. If two or three deals slip simultaneously, the runway math changes.
\nWe are managing this by targeting smaller health systems first (shorter cycles), building a pipeline of 8–10 opportunities so that 2–3 slipping does not derail the milestone plan, and pursuing retrospective data agreements (which have shorter procurement cycles than prospective pilots) to generate outcome data in parallel. The 18–20 month runway is specifically sized to absorb procurement delays while still hitting the Series A milestones.
\n" } ] }