One Billion Health Records Flow Through TEFCA, Driving Metriport Hiring

Hiring Surge: Scaling to Millions of Patient Records

Metriport added six roles in the past week (Account Executive, Staff Data Engineer, Staff Software Engineer, Senior Data Engineer, Senior Software Engineer, and Technical Account Manager) with salary bands reaching $260,000, per the company's job board. The federal interoperability mandate now moves more than one billion records through TEFCA networks, per HHS data from June 2026.

The company's Medical API connects to the largest health information exchanges in the country and covers 300 million patients through a single open-source interface, a figure published on its GitHub repository (20,135 commits as of June 2026). The documentation describes the work as "standardizing, de-duplicating, consolidating, and hydrating data with medical code crosswalking."

Metriport raised a $2.4 million seed round at a $20 million valuation in December 2022, backed by Y Combinator, Triple Impact Capital, Nueterra Capital, and a syndicate of angels. The round closed before the Medical API launched; the Health Device API (wearables and RPM data) shipped first. Founders told TechCrunch the Medical API was the "next step" and customer demand was already lining up. Two years later, the job board shows the team converting that demand into headcount. Roles are concentrated in San Francisco, on-site, with bands that assume seniority.

Customers validate the bet. Kivo Health's CEO, Victor Sadauskas, said Metriport's engineering and onboarding support automated patient-record retrieval and cut clinical note access to seconds. Tono Health's co-founder and CMO, Dr. Mondana Ghias, cited depth of coverage, data quality, and normalization as reasons the integration proved "invaluable." Both build on Metriport's open-source core.

Metriport positions itself as the "Plaid for healthcare" — a universal API layer that abstracts the fragmentation underneath. Plaid's model relied on screen scraping and bank partnerships; Metriport's relies on HIEs mandated by federal rule (TEFCA, covered next). The salary bands reflect the asymmetry: a wrong medication list in fintech means a declined transaction; in healthcare, it means a contraindicated prescription. The board shows the company pays for engineers who treat data quality as a correctness problem.

TEFCA's Real-Time Pipeline Problem

The Trusted Exchange Framework and Common Agreement went from policy framework to production network in December 2023, when the first Qualified Health Information Networks were designated and health data began flowing among them. By February 2026, HHS reported nearly 500 million health records exchanged; four months later that figure crossed one billion. More than 23,000 organizations (QHINs, Participants, and Subparticipants) now represent over 100,000 unique connections to clinicians, hospitals, clinics, post-acute facilities, and public health authorities, per the RCE. The Common Agreement Version 2.1, released in November 2024, and the Draft QHIN Technical Framework Version 2.1, opened for feedback in June 2025, both tighten performance measurement and patient-identity resolution requirements that cascade down to every participant.

For a platform like Metriport, which aggregates records from the largest HIEs through a single open-source API, TEFCA's scale creates a concrete engineering problem: the exchange purposes defined in the framework — Treatment, Payment, Health Care Operations, Public Health, Government Benefits Determination, and Individual Access Services — demand low latency when a clinician pulls a patient's chart at the point of care. The QHIN Technical Framework specifies authentication, patient identity resolution, and performance measurement as mandatory components. Metriport's pipeline must normalize FHIR, C-CDA, and PDF payloads from disparate HIEs, de-duplicate and hydrate them, and surface a unified record in seconds — not the hours or days that point-to-point connections historically required.

The Sequoia Project, serving as the Recognized Coordinating Entity under a five-year ONC contract awarded in August 2023, had made clear that the Common Agreement's baseline requirements would continue to evolve. ONC and the RCE plan to expand permissible exchange purposes over time based on market needs raised by the QHIN Governing Council. The framework also prohibits throttling, limiting data elements, or imposing burdensome testing requirements that would impede exchange: constraints that rule out batch-oriented architectures in favor of streaming, event-driven pipelines.

Metriport's recent hiring for Staff and Senior Data Engineers maps to this workload: building connectors that ingest from TEFCA-aligned QHINs, implementing the patient-matching algorithms the QTF mandates, and operating the observability stack that proves compliance with performance metrics. The open-source approach means those connectors and normalization routines are developed in public repositories, subject to scrutiny from both the RCE and the QHINs themselves.

HHS frames TEFCA as decreasing costs for providers by eliminating the need to join multiple networks or create one-off, point-to-point connections. Metriport's product pitch mirrors that language: one API, the major HIEs, seconds not days. Every new QHIN onboarding, every Common Agreement revision, and every added exchange purpose multiplies the permutation space the pipelines must handle. The hiring surge is about parallelizing the work of mapping TEFCA's expanding technical surface area into reliable, auditable code paths that can sustain billion-record throughput.

Clinical Workflow: From Days to Seconds

The bottleneck in modern clinical workflows isn't the visit itself — it's the scramble before the patient walks in. On average, a doctor spends 16 minutes with a patient, diagnosing and taking notes, per a Medscape survey cited in TechCrunch. When the chart arrives late or incomplete, those minutes get eaten by hunting for lab results, medication lists, or discharge summaries that should have been there already.

Metriport targets that pre-visit window directly. The company frames the value proposition as "Get comprehensive patient medical data in seconds, not weeks", a claim repeated across its website and GitHub repository. Traditional record requests routed through fax, portal logins, or manual HIE queries routinely take days. Metriport aggregates from the largest HIEs in the country via that API, delivering FHIR R4, C-CDA, and PDF formats in real time.

The operational shift appears in production dependencies, not pilots. Amazon One Medical, a Metriport customer, operates in the digital-first primary care segment where pre-visit data latency directly undermines the promise of AI-augmented visits. Metriport's real-time pipelines and notification feeds are designed to close that gap.

The implication extends beyond primary care. Specialty practices receiving referrals, behavioral health teams coordinating with medical providers, and value-based care organizations tracking readmission risk all depend on the same data arriving before the clinical decision point. Metriport's "Integrate in hours, not months" claim (backed by a single API covering multiple HIEs) reduces the integration surface that historically forced each new data source into a custom project.

No independent third-party study quantifies the time savings across Metriport's customer base. The evidence consists of the company's published benchmarks, its GitHub commit velocity, and the qualitative testimonials from named customers like Kivo and Tono. When the data retrieval layer shifts from batch-oriented, multi-vendor workflows to a unified real-time API, the clinician's prep time collapses from hours of chasing records to a single query returned before the exam room door opens.

Legacy Vendors React: Litigation, Silence, and the Open-Source Threat

Epic Systems filed a federal complaint in January 2026 in a California district court. The EHR giant and several large health systems accused a network of health data intermediaries and downstream companies of exploiting national interoperability frameworks (TEFCA and Carequality among them) to access patient records under false pretenses and then monetize that data for purposes unrelated to patient care. The suit alleges nearly 300,000 patients had their records improperly accessed. Epic's filing frames the defendants as abusing the "trusted exchange" architecture the federal government built to enable legitimate care coordination. The complaint does not name Metriport, but the described behavior (aggregating records from HIEs via API and making them available to digital-health companies) maps closely to Metriport's stated model: a single open-source API pulling from the largest clinical data networks in the country.

STAT News reported Epic saying it had "uncovered rot in patient record sharing" and pressed HHS to act. The message to Washington: the current TEFCA governance allows bad actors to masquerade as treatment-purpose participants.

Meanwhile, Epic has not shut its own doors to outside developers. The company maintains a SMART on FHIR R4 implementation and an app marketplace where third parties can embed inside the Epic workflow. Metriport publishes an Epic embedded app that lets clinicians initiate network queries and view Metriport-aggregated records without leaving the Epic interface. That integration exists because Epic's platform permits it: a reminder that the same vendor suing intermediaries also provides the rails those intermediaries ride.

Cerner, now owned by Oracle, has responded differently. Oracle's health business has been cutting staff. The VA's EHR rollout (the largest Cerner/Oracle deployment) continues, but the program's history of delays and cost overruns has made it a cautionary tale rather than a competitive moat. Oracle has not filed a counterpart lawsuit, and its public posture emphasizes cloud infrastructure and AI services layered atop Cerner's data rather than legal defense of data boundaries.

Metriport's open-source posture has forced a conversation that proprietary FHIR-api vendors would rather avoid. The company's public repository (759 stars, 107 forks, and more than 20,000 commits as of June 2026) is the product surface that customers, security researchers, and rival engineers can inspect line by line. Metriport's blog frames this as a structural advantage: "Open source code doesn't build trust with perfect matching, but with transparency: everyone can see how the decisions are made." The trade-off between match rates and potential errors is published, not buried in a sales deck.

Metriport aggregates the largest HIEs into a single API covering 300 million-plus patients, providing those formats in seconds. Its marketing promises integration in hours, not months, a direct shot at the months-long onboarding cycles typical of proprietary middleware. When a digital-health company can self-host Metriport's code and avoid vendor lock-in entirely, the switching cost that protects incumbent FHIR platforms collapses.

Epic has signaled discomfort through litigation and public statements rather than product announcements. A 2024 Forbes investigation detailed a "bombshell Epic lawsuit" over alleged misuse of patient records, and STAT reported Epic claiming it "that claim" while pressing HHS to act. Those moves read less like quality enforcement than like a gatekeeper defending its toll booth.

Customer testimony reinforces the threat perception. Amazon One Medical, Brightside Health, and Circle Medical (each a scale digital-health operator and Metriport customer) cite the platform's depth of coverage, normalization quality, and speed of integration. Dr. Mondana Ghias of Tono Health said: "Integrating Metriport's medical record summaries directly into our EHR has proven invaluable. Metriport goes above and beyond with those qualities." Those are production dependencies on open-source infrastructure.

What the public record lacks is on-the-record commentary from proprietary FHIR vendors explicitly naming Metriport as a competitive threat. No major middleware provider has issued a white paper or earnings-call remark addressing the open-source aggregator model. Until a proprietary vendor opens its own matching logic or matches Metriport's self-hosting guarantee, the competitive dynamic remains asymmetric — one side publishes its trade-offs, the other asks for trust.

The YC Playbook in Action

Metriport's founding team carries the Y Combinator stamp directly: Colin Elsinga and Dima Goncharov went through the Summer 2022 batch, and the company's job postings still lead with that credential: "a tight-knit, high-performing team of mostly former founders (including two YC alumni)." Goncharov's background reads like a YC archetype: aerospace systems, highly available cloud databases, then a pivot to consumer health. Elsinga built consumer health apps before co-founding Metriport. Both are software engineers first, operators second.

The company is 26 people, yet 13 roles are open — nearly all in engineering. The salary bands cluster tightly around market rates for senior ICs in San Francisco, but the equity component and the "engineering-heavy" label suggest the compensation conversation starts with technical leverage.

The culture language in the job posts is unvarnished: "We push hard — founders work six days a week from our SF office — but give everyone freedom to craft their schedule." "We measure output and we're committed to sustainable intensity." "Low ego, high ownership. You ask 'Why can't this be done in 3 days?' instead of 3 weeks." These are filters. A candidate who wants process documentation before writing code will self-select out. A candidate who has shipped a data pipeline from scratch in a weekend will lean in.

The YC model (compress timelines, default to action, iterate in public) maps directly onto the open-source strategy. The same repository that hosts Metriport's FHIR normalization logic also hosts its hiring signal: transparency, no vendor lock-in, community contributions welcome.

Quarterly off-sites and unlimited PTO appear in the benefits list, but the real retention mechanism is autonomy: "Operate with minimal bureaucracy and high autonomy, hire based on competence, not prestige." That sentence attracts engineers who have been slowed by approval chains at larger health-tech vendors, and repels managers who need hierarchy to function. The board data shows a team scaling without a dedicated engineering management layer in the open roles — just ICs and a Design Engineer. That flatness is a direct inheritance from the YC playbook: keep the org chart shallow until the product forces structure.

The open-source mandate reinforces the hiring profile. Engineers who publish their own libraries, contribute to FHIR tooling, or run self-hosted infra in their spare time are already speaking Metriport's language. The company's site frames it as "Rely on our immortal code, not our business" — a recruiting pitch disguised as a customer promise. Candidates who read the GitHub repo before the interview come in knowing the stack, the failure modes, and the community norms.

The tension is visible in the numbers: 26 people, 13 open roles, six-day founder weeks. The YC alumni culture that accelerated Metriport to multi-million ARR and 100-plus customers now has to scale without diluting the very intensity that got them here. The hiring surge is the test. Every new Staff Data Engineer hired at $260k with equity is a bet that the "sustainable intensity" claim holds. If it doesn't, the open-source advantage (community velocity, external audits, no black boxes) erodes with the culture that produced it.

Scope Note: What This Story Doesn't Cover

This article tracks Metriport's hiring surge for data engineers and the federal interoperability mandates driving it. It does not cover the company's financial performance, fundraising history, or detailed product roadmap beyond what bears directly on that hiring narrative.

The financial picture in public sources is fragmented and, in places, contradictory.

Those salary bands and the hiring velocity are the only compensation and headcount figures this article treats as current and reliable.

On the product side, research notes a roadmap that includes expanding healthcare network connectivity, EHR applications, clinical summaries, patient monitoring, analytics, and data-contribution workflows, language drawn from a Biobase profile that explicitly states Metriport is not a therapeutics company and has no drug-development pipeline. Fierce Healthcare reported in 2023 that the company planned to approach additional health information exchanges, including state-level HIEs.

The omission is deliberate. The hiring surge (12 employees as of December 2024 per GetLatka, 11–20 per Prospeo, and the live board showing active recruitment for senior and staff-level data engineers) is the observable signal that the federal TEFCA mandate and customer demand from organizations like Amazon One Medical are translating into engineering capacity. Financial metrics and product feature lists are lagging or forward-looking indicators; they do not change the fact that Metriport is hiring data engineers now, at scale, to build the pipes that Epic's lawyers and TEFCA's rules are fighting over.

Working in frontier tech? Zero G Talent tracks the openings: see every open Metriport role, browse frontier tech jobs, the companies hiring, and the people building the field.