Table of Contents

Hubble Investment Report

Hubble Investment Report

Category: Healthcare data infrastructure / developer API (health IT)

Company Stage: Pre-seed (YC S26); founded 2026

Founder or Founders: Prabha Dublish (CEO), Aaron Leon (CTO)

Headquarters: Seattle, WA

Funding: Y Combinator standard deal ($500K); no other funding publicly disclosed

Business Model: B2B API platform (usage-based retrieval; pricing not publicly disclosed)

Product Hunt Launch Date: August 18, 2026

Report Date: August 21, 2026

Investment Metric	Assessment
Venture Potential	54/100
Unicorn Path	Conditional
Valuation Attractiveness	Not Assessable
Evidence Confidence	45/100
Final Decision	Watch

Executive Summary

Hubble (hubble.ai) is a medical-records retrieval API: a patient verifies their identity once, and Hubble assembles their records across providers and payers, returning one normalized record. Where network APIs return nothing, it falls back to browser agents that work provider portals and voice agents that call records departments, all under the patient’s HIPAA individual right of access (Product Hunt; patient-mediated access page).

The product serves developers and organizations that need patient records: digital health and AI builders, legal (personal injury, mass tort, disability), and life sciences research, per the company’s Launch YC post and founder profiles.

The strongest positive signal is founder-market fit. CEO Prabha Dublish was the first product hire at Grow Therapy (which reached a $3B valuation in March 2026) and CTO Aaron Leon was a tech lead at Amazon One Medical working on its in-house EHR and AI tooling, per the YC company page. Both have personally operated the systems Hubble now integrates with.

The central concern is that there is no verifiable commercial traction. The company is roughly three months old, launched publicly this week, and discloses no revenue, customers, pricing, or retention data. Claims such as “70,000+ connected systems” and “live in production” are company-reported and unverified.

The decision is Watch. The market is real and the founders are credible, but the venture case currently rests on a launch, not on evidence of paying customers or durable differentiation against well-funded incumbents.

Product Overview

The customer problem is genuine: patient records are fragmented across EHRs, HIEs, payer systems, and fax-based release-of-information departments, and existing network APIs typically return only the portal-accessible subset. Hubble’s workflow: the patient verifies identity to NIST IAL2 (government ID plus liveness check), authorizes retrieval under HIPAA right of access (45 CFR 164.524), and Hubble pulls clinical data and documents from EHRs directly and via HIE networks, plus payer claims where authorized, normalizing everything into one record with per-field source attribution and an audit trail. Specially protected records (42 CFR Part 2 substance-use, psychotherapy notes, HIV/genetic, minors) follow a separate consent path (patient-mediated access page).

Core features: patient-mediated retrieval (live), provider-mediated access in beta (eligibility, prior auth, scheduling, claims), voice and fax agents for non-API sources, PDF export, and an MCP server for AI agents (site; data infrastructure page). Pricing is not publicly disclosed; Product Hunt lists “Free Options” and a 7-day free launch offer. The product replaces manual fax/phone retrieval, per-provider integration projects, and network-only APIs such as Metriport or Particle Health.

Founder and Team Assessment

Both founders are identity-verified through LinkedIn and the YC directory. Dublish’s LinkedIn confirms “Co-Founder @ Hubble (YC S26),” Seattle, starting May 2026, and prior roles as Grow Therapy’s first product hire and Meta PM/PMM. Leon’s background (One Medical tech lead, 50+ engineers, Amazon Care data platform) is stated on the YC page and company site; the One Medical EHR (“1Life”) context is corroborated by an AWS case study, though his specific role is company-reported. The founders have known each other for over a decade and previously built a nonprofit together (LinkedIn post).

Team size is two, per YC. No prior exits. Key-person risk is high by construction. Neither founder has run a sales-driven healthcare infrastructure company, so commercial capability is unproven despite strong product credentials.

Founder Assessment: Exceptionally relevant technical and product experience for this exact problem, but commercial execution and team-building remain unproven.

Market Opportunity

The initial customer is narrow: US digital-health and legal-tech teams that need complete patient records and cannot get them through network-only APIs. Bottom-up: the US has tens of millions of record-retrieval events annually across care coordination, legal, and insurance workflows; the incumbent Datavant/Ciox combination reported roughly $700M in combined annual revenue and a $7B merger valuation in 2021, demonstrating that records retrieval alone supports large revenue pools. Adjacent expansion includes provider-mediated workflows (prior auth, eligibility), payer integrations, and research data — each a larger budget than retrieval itself. Willingness to pay is established in legal/insurance retrieval (per-record fees) and emerging among AI builders. The realistic addressable market can support venture-scale revenue, but Hubble must win share from entrenched, capitalized incumbents rather than create a new category.

Traction and Growth Signals

Available evidence: Product Hunt “Launch of the Day” on August 18, 2026, with 327 product followers and active founder engagement in comments (Product Hunt; awards); a Launch YC post (August 11, 2026); inclusion in the YC S26 batch; and a Product Hunt newsletter feature (archive). Company-reported but unverified: “live in production,” “70,000+ connected systems,” 40+ tools live (site; Extruct profile).

There is no public evidence of revenue, paying customers, retention, or usage volume. The most important missing metrics are paying customer count, completed retrieval volume, retrieval success rate, and revenue. This is launch attention, not sustained traction — appropriate for a company this young, but not evidence of product-market fit.

Traction Assessment: Promising launch reception, commercially unverified.

Competitive Position

Direct competitors include network-based record APIs: Particle Health ($39.3M raised), Health Gorilla ($50M Series C), open-source Metriport, and patient-consent-first Flexpa (~$11M raised from Founders Fund, Khosla, Ribbit and others). Incumbent retrieval giants (Datavant/Ciox, MRO, Verisma, Ontellus) own legal/insurance channels. Adjacent AI automation players like Tennr ($101M Series C at a $605M valuation) prove demand for voice/fax automation in healthcare and could expand into retrieval.

Hubble’s differentiation is the multi-modal fallback (API → browser agent → voice agent) under patient right of access, plus source-attributed normalization. However, the legal basis is available to anyone, and the agent layer is replicable. If Epic, a large HIE, or Datavant shipped equivalent patient-mediated retrieval with voice fallback within six months, Hubble’s residual advantage would be developer experience and retrieval completeness — meaningful but not yet proven. Improving native APIs (TEFCA, CMS-9115 payer APIs, ONC (g)(10)) could also shrink the “long tail” that justifies the product.

Defensibility Assessment: Low-to-medium today, with a credible path to medium via operational edge-case data and retrieval-success rates.

Business Model and Economics

The model is a B2B API, almost certainly usage-based per retrieval plus platform fees, but pricing is not publicly disclosed, so ACV, gross margin, and conversion are not assessable. Structurally, electronic API pulls should carry software-like margins, while voice/browser retrievals incur telephony, inference, identity-verification (IAL2), and human-escalation costs — the founders acknowledged human escalation paths in PH comments. The key economic question is whether blended per-retrieval revenue exceeds fully loaded retrieval cost at scale, and whether customers pay subscription minimums. Expansion revenue potential exists via provider-mediated workflows and payer data. All unit-economics claims require verification.

Unicorn Path

Assume an 8–12x forward-revenue multiple, appropriate for high-growth health-data infrastructure with usage-based revenue (consistent with Tennr’s ~$605M valuation on rapidly scaling but unprofitable revenue). Required revenue for $1B: roughly $85–125M ARR. As an analyst assumption (not company data), if blended net revenue per completed retrieval is $20–40, Hubble would need roughly 2.5–6M completed retrievals annually, or a smaller base of enterprise contracts at $250K+ ACV — for context, Datavant/Ciox reached ~$700M revenue only after consolidation at national scale. This requires expanding from retrieval into a broader records platform (provider-mediated read/write, payer workflows, research data), repeatable enterprise distribution, and a much larger team. Achievable in this market, but only with successful business-model expansion beyond the launch product.

Unicorn Path: Conditional

Valuation Assessment

Known funding: YC S26 participation, which under YC’s standard deal means $500K ($125K for a fixed 7% plus a $375K uncapped MFN SAFE). No priced round, SAFE cap beyond YC, revenue, or growth data is disclosed. Comparable financings (Flexpa’s $9M seed; Particle’s early rounds) suggest seed-stage health-data APIs price in the $10–25M post-money range, but Hubble’s terms are unknown.

Valuation Attractiveness: Not Assessable. Required: current ARR, retrieval volume and success rate, gross margin, burn, round size, and post-money or SAFE cap.

Key Risks

No verifiable commercial traction — revenue, customers, and pricing all undisclosed.
Competitive replication: funded incumbents (Particle, Health Gorilla, Flexpa, Datavant) can add agent-based fallback.
Native API improvement (TEFCA, CMS-9115, (g)(10)) erodes the long-tail gap the product exploits.
Voice-agent reliability and compliance risk when providers challenge automated callers.
Unit economics: IAL2 verification, telephony, inference, and human escalation may compress margins.
Regulatory exposure handling sensitive records (HIPAA, 42 CFR Part 2, state laws).
Two-person team with acute key-person risk and no demonstrated hiring.
Long healthcare enterprise sales cycles versus a developer-style GTM.
Dependence on EHR/HIE cooperation and portal terms of service.
Willingness to pay unverified in the absence of any pricing signal.

Final Assessment

Venture Potential: 54/100

Category	Score
Market Size and Expansion Potential	15/20
Traction and Growth Evidence	4/20
Founder and Team	11/15
Product Strength	7/10
Distribution Potential	8/15
Business Model and Economics	5/10
Defensibility	4/10
Total	54/100

Strongest elements: a large, proven market and founders with rare, directly relevant experience. Weakest: zero disclosed commercial traction and thin structural defensibility.

Evidence Confidence: 45/100

Verified: founder identities, YC S26 backing, Seattle base, two-person team, product positioning, launch results. Company-reported: “live in production,” 70,000+ connected systems, prior-employer impact claims. Unavailable: revenue, customers, pricing, margins, retention, funding beyond YC.

Final Decision: Watch

The company is days past public launch with strong founders in a real market, but every commercial question is open. This fits Watch precisely: promising product and positive community reception, with revenue, retention, and the venture-scale path all unverified.

Upgrade Conditions

10+ verified paying customers or $250K+ ARR within two quarters
Documented retrieval success/completeness rates versus network-only APIs
Published pricing with evidence of paid conversion
A priced seed round with credible healthcare-infrastructure investors
Blended gross margin above 60% including voice-agent costs

Downgrade Conditions

No paying customers announced within 6–9 months
A funded incumbent shipping equivalent voice/browser retrieval
Provider backlash or legal challenges to voice-agent retrieval
Founder departure or product inactivity
Unit economics requiring sustained human escalation per retrieval

Questions for Further Diligence

Current MRR and the number of paying versus trial customers?
Completed retrieval volume to date, and average turnaround time per record?
Retrieval success rate by channel (API vs. browser vs. voice), and completeness versus network-only baselines?
Fully loaded cost per retrieval, including IAL2 verification, telephony, inference, and human escalation?
What pricing model are early customers accepting, and at what ACV?
How do providers respond to voice agents today, and what is the escalation rate to humans?
What is the legal review behind right-of-access retrieval at scale, including 42 CFR Part 2 handling?
Which customer segment (legal, digital health, research) is converting fastest, and why?
What stops Particle, Flexpa, or Datavant from replicating the agent layer?
Current burn, runway, and terms/cap of any active or planned raise?
Hiring plan for the next 12 months, and which roles first?
What usage or edge-case data accumulates as a proprietary asset over time?