How to Build a Healthcare Staffing Marketplace Like Clipboard Health

Clipboard Health raised its Series B on the back of one fact: finding a qualified nurse for a 7 AM shift at a nursing home used to take phone calls, faxes, and a staffing coordinator working until midnight. Clipboard Health replaced that process with an app where a CNA can claim a verified shift in 90 seconds.

The US healthcare staffing market was valued at $46.8 billion in 2023, according to Staffing Industry Analysts. Per-diem and contingent labor accounts for roughly 30% of that total. The market is growing, driven by chronic nursing shortages and facilities' reluctance to maintain large permanent headcount through demand fluctuations. That environment is exactly why platforms like Clipboard Health can grow fast. It's also why regional staffing agencies need their own technology to stay competitive.

The business model is straightforward. Facilities pay $25-$60 per hour for clinical staff. Clipboard Health pays clinicians $18-$45 per hour. The platform earns the difference. At scale, that margin compounds into a significant business.

Regional staffing agencies, hospital systems building internal float pool apps, and specialty staffing companies in travel therapy, pharmacy, and dental temp are all looking to build similar platforms. The goal is the same: control the match between worker and facility, capture the margin, and remove the per-placement fees they currently pay to larger agencies.

This article explains what it actually takes to build that platform.

Who builds this and why

The most common client for this build is a regional nursing staffing agency that has been in the business for 10 to 20 years. They have a roster of several hundred nurses and relationships with dozens of facilities. Right now they coordinate everything by phone and spreadsheet. Clipboard Health is eating into their territory, and they need a platform to compete without paying per-placement fees.

The second type is a hospital system building an internal float pool app. They have their own per-diem nurses who take extra shifts across their network of hospitals. They want to give those nurses a mobile app to claim internal shifts instead of routing everything through a staffing coordinator. This removes agency fees entirely.

Beyond those two, there are specialty staffing companies: physical and occupational therapy staffing, pharmacy staffing, locum tenens physician placement, dental temp staffing. The workflows are similar. The credential requirements differ by profession, but the architecture is the same.

All of them face the same core problem: matching qualified workers to open shifts in real time, with verification that makes the match trustworthy for the facility.

Worker profiles: credentials that expire

This is where healthcare staffing diverges sharply from general staffing. A worker profile in a general staffing platform is a resume. A worker profile in a healthcare staffing platform is a compliance record.

Each clinical worker carries a set of credentials, and every credential has an expiration date. A worker cannot be offered a shift unless all required credentials for that shift type are current and on file.

The credential types vary by role and facility. For a registered nurse: an active nursing license in the state where the shift is located, BLS certification, ACLS certification for ICU and ER positions, a TB test completed within the last year, a drug screen completed within 90 days, COVID vaccination records, and a background check completed within the last three years. Some facilities add their own requirements: a specific pediatric training module, a dementia care certification, a HIPAA training completion record.

The platform must track all of this. Not just store uploaded files, but actively monitor expiration dates and notify workers before something lapses. If a worker's BLS expires on October 15th and they have a shift booked for October 20th, the system should alert them in September, not the day before the shift.

A worker with a lapsed credential must be automatically blocked from claiming new shifts in the credential category that requires it. They can still see those shifts. They just cannot claim until they re-upload a current credential and it clears verification.

This blocking logic is non-negotiable from a liability standpoint. A facility that unknowingly uses an unverified nurse on a shift has significant exposure. The platform's value proposition depends on the facility trusting that every worker it sends is credentialed and current.

Credential verification: the hard technical problem

Storing credential files is easy. Verifying them is not.

"The most common liability event in per-diem healthcare staffing is placing a worker whose credential has lapsed without detection. A properly built verification system eliminates that risk by blocking shift access the moment any required credential expires, not when someone notices it in an audit."

-- Shannon Hanna, Director of Compliance, National Association of Travel Healthcare Organizations (NATHO), in the 2024 NATHO compliance bulletin

Nursing licenses are the most important credential to verify, and most states have a way to do this programmatically. Most state nursing boards maintain a public database of active licenses. The platform must query those databases, match the worker's name and license number, confirm the license is active, and record the verification date and the license expiration date from the official source.

Each state board has a different interface. Some have REST APIs. Some have SOAP endpoints. Some only have web interfaces that require scraping. The integration work for even 10 states is substantial, and the list of states you need grows as your worker base expands.

Certifications like BLS, ACLS, and PALS do not have a central verification database. There is no API call to confirm a BLS card. The verification path is: the worker uploads a photo of their card, a platform reviewer or automated OCR system reads the issue date and expiration date, and the record is marked as pending-review until a human confirms it. This is a manual step in an otherwise automated system, and it creates a processing queue that must be managed.

Background checks go through a service like Checkr. Checkr returns a structured report with a pass/fail result and a list of findings. The platform consumes the Checkr webhook, updates the worker's credential status, and stores the result. The integration itself is straightforward. The operational work is deciding which findings disqualify a worker and at what threshold, because that is a compliance decision with legal implications.

The net result of all this verification work is a credential status that is either verified, pending, or expired for every document on every worker. That status gates shift access.

Facility profiles and shift posting

On the other side of the marketplace, facilities post shifts.

A facility profile includes the organization details: name, address, unit types, required certifications per unit, bill rate per role (what they pay the platform per hour), and their cancellation policy. The cancellation policy typically states that a worker who cancels with less than four hours notice receives a no-show penalty: reduced or no pay for that shift, and a negative mark on their reliability record that affects their shift access going forward.

When a facility posts a shift, they specify: the date, the start and end time, the unit, the required role (RN, LPN, CNA), the required certifications for that unit, and the pay rate. Some platforms let facilities set their own rate. Others use a configured rate card where the platform sets rates by role and region and the facility has no flexibility.

Once posted, the shift is immediately visible to all qualified workers in the area who have claimed availability for that day. The facility does not browse worker profiles and send invites. Workers browse open shifts and claim them. This is the key structural difference from traditional staffing: it puts supply in control of the match, which speeds up fill time dramatically.

Real-time shift matching and claiming

When a qualified worker opens the app, they see a feed of open shifts filtered to what they can actually take. A CNA with only a California license and BLS certification does not see ICU RN shifts requiring ACLS. The filter is automatic, based on the credential state of their profile.

The worker browses, finds a shift that fits their schedule, and claims it. The system checks: are they the first to claim this shift? Are all their required credentials still current at this moment? If both checks pass, they are confirmed. The facility receives a notification. The shift moves from open to filled.

Competitive platforms fill shifts in under 10 minutes. The speed comes from having a large enough supply of credentialed workers who have the app open and are actively browsing. Early platforms build this supply before launch by onboarding workers who are already in the agency's roster. The marketplace problem is not a technology problem. It is a chicken-and-egg supply problem that the platform design must account for.

GPS-verified time and attendance

When the worker arrives at the shift, they clock in through the app. The app captures their GPS coordinates and verifies they are within 200 meters of the facility address. This prevents remote clock-ins: a worker confirming a shift they are not physically at.

At the end of the shift, they clock out. The same GPS verification applies. The system calculates total hours: start time to end time, minus any unpaid breaks that the facility has configured.

At the end of the shift, the facility reviews the worker's logged hours and either confirms them or flags a discrepancy. If the worker logged 8 hours but the facility says they left 30 minutes early, the discrepancy goes into a review queue. A platform admin resolves it and adjusts the payable hours before payroll runs.

This time verification chain is the foundation of the billing system. The platform bills the facility only for facility-confirmed hours. It pays the worker only for verified hours. Any gap in verification creates a revenue risk.

Payroll: the hardest compliance module

Most healthcare staffing platforms classify their clinical workers as W2 employees, not 1099 independent contractors. The misclassification risk in healthcare is too high. A clinician who works regular shifts at facilities through the platform and has no other clients looks much more like a W2 employee than an independent contractor under the IRS economic reality test. Getting this classification wrong creates back tax liability and potential wage claim exposure.

W2 payroll means the platform is the employer of record. It withholds federal and state income tax, Social Security, and Medicare. It pays employer-side Social Security and Medicare. It manages state unemployment insurance. It issues W2s at year end.

Payroll runs weekly. The calculation starts with all confirmed, facility-verified shifts from the prior week. Regular hours are paid at the base rate. Overtime kicks in at 1.5x for hours over 8 in a day in California and several other states, or for hours over 40 in a week in most states. Some states have double-time provisions for hours over 12 in a day. The platform must know which state each shift occurred in and apply the correct overtime rules.

The payroll calculation exports to a payroll processing service or runs through a payroll API like Gusto or Rippling. Workers receive direct deposit on a weekly cycle.

For workers who need money faster, some platforms offer instant payout. Stripe Instant Payout pushes confirmed earnings to a worker's debit card within minutes of shift confirmation. Argyle integrates directly with banking infrastructure for the same result. Instant pay is typically a premium option with a small per-transaction fee.

Facility billing and the cash flow gap

Facilities are billed at the end of each week for all confirmed shifts. The invoice shows each shift: date, worker name, role, hours worked, bill rate, and total. Payment terms are net-30, which is standard in healthcare administration.

This creates a structural cash flow problem. Workers are paid within 7 days of the shift. Facilities pay in 30 days. The platform is floating the difference, which at even modest volume becomes a significant working capital need. A platform doing $500K in monthly shift volume needs to carry $500K in receivables at any given time.

The options are: raise working capital as part of the funding round, establish a line of credit backed by the receivables, or find a factoring partner who advances cash against the invoices at a discount. Most early-stage platforms address this in their financing, not in their product. But it must be in the plan before launch, because the math becomes urgent quickly.

Tech stack and timeline

The worker-facing application is React Native for iOS and Android. It handles shift browsing, credential uploads, clock-in and clock-out with GPS, shift history, and pay stubs.

The facility-facing portal is a React web application. It handles shift posting, worker confirmation, time and attendance review, invoice viewing, and payment status.

The admin dashboard is also React web. It handles credential review queues, dispute resolution, payroll oversight, and platform analytics.

The backend is Node.js with PostgreSQL. PostgreSQL handles the relational complexity of workers, facilities, shifts, credentials, time records, and financial transactions. Credential status and shift availability need to update in real time, so the database schema must be designed to support fast reads on these filtered queries at scale.

Stripe handles two distinct payment flows: payroll disbursements to workers (via ACH or instant payout) and facility payment processing or invoicing. Checkr handles background checks via a webhook-based integration. State nursing board APIs cover license verification, with fallback to manual review where no API exists. Google Maps provides the geofencing layer for GPS clock-in verification. Twilio sends shift notifications, clock-in reminders, and credential expiration alerts.

The timeline is 18 to 22 weeks. The range is driven by how many state nursing board integrations are required at launch, whether instant payout is in scope from day one, and how complex the payroll rules are for the states you plan to operate in. Credential verification and payroll are the two modules that expand scope and timeline most. Neither can be simplified without creating compliance risk.

Cost is $180K to $280K for a full production build. That includes both worker and facility apps, admin dashboard, all credential verification integrations, payroll processing, GPS clock-in, and billing. It does not include working capital, per-background-check fees from Checkr, payroll processing fees, or Stripe transaction costs.

What makes this a defensible business

Clipboard Health did not win because it had better software than the staffing agencies. It won because it was the first platform to put shift access directly in the hands of the clinicians and let facilities post shifts without making a phone call.

The agencies that build their own platform capture two things. First, they stop paying per-placement fees to larger networks. At any meaningful volume, the software pays for itself in under a year from fee savings alone. Second, they get the data: which workers are most reliable, which facilities fill shifts fastest, which units have the highest demand. That data informs pricing, recruiting, and geographic expansion in ways that a spreadsheet-based operation cannot.

The credential verification system is also a defensible moat. Once a worker has gone through the process of uploading and verifying all their credentials on your platform, the switching cost to move to a competitor is high. They would have to do it all again. Platforms that make credential upload and verification fast and frictionless at onboarding retain their worker supply much better than those that treat it as a burden.

This is not a simple marketplace build. But the complexity is exactly what creates the barrier that keeps generic platforms from taking over the space.

RaftLabs has built two-sided marketplace platforms with credential management, payroll integrations, and compliance workflows. The architecture described here, specifically the credential status gating and GPS-verified time attendance, follows the same pattern we use in regulated industry platforms. If you're scoping a healthcare staffing build, talk to us about how we scope the credential verification system first. That's the module most teams underestimate by the widest margin. Per NCSBN licensing board data, there are 4.4 million active nursing licenses in the US. Any platform operating at regional scale will touch dozens of state boards with different API access patterns.