Custom Software Development
Systems built around how your business actually works.
Healthcare · 2025
Cutting patient intake from 22 minutes to under 6
A specialty medical group operating fourteen locations was losing clinical capacity to paper intake and manual insurance verification. We built a HIPAA-compliant digital intake system integrated with their EHR.
- Client
- Multi-site specialty group
- Duration
- 5 months
- Services
- 4 practices involved
Placeholder content. These engagement details illustrate structure and formatting. They are not reported client results and will be replaced with verified client stories before this site goes live.
Results
- Average intake processing time
- 22 → 5.8 minAverage intake processing time
- Insurance-related claim denials
- -71%Insurance-related claim denials
- Patients completing intake before arrival
- 81%Patients completing intake before arrival
- Locations live within eleven weeks
- 14Locations live within eleven weeks
The challenge
What was actually going wrong
Patients arrived twenty minutes early to complete paper forms that front desk staff then keyed into the EHR. Between transcription, insurance verification and correcting illegible entries, each new patient consumed roughly twenty-two minutes of staff time before any clinical work began.
The error rate compounded the problem. Insurance information entered incorrectly produced claim denials weeks later, and the billing team spent significant time on rework that originated at the front desk.
The group had evaluated the EHR vendor's own intake module and found it required patients to create an account before completing forms — a step that testing showed a substantial share of older patients abandoned.
Our approach
How we worked the problem
- 01
Remove the account requirement
Intake links are single-use, time-limited and tied to the appointment. Patients complete forms without creating credentials, which removed the largest abandonment cause in the vendor product.
- 02
Design for the actual patient population
The specialty skews older. We tested with patients over sixty-five, which drove larger type, fewer questions per screen, obvious progress indication and a phone-friendly layout with generous tap targets.
- 03
Verify insurance automatically
Eligibility checks run on submission against the clearinghouse, so coverage problems surface before the visit rather than after the claim. Failures route to staff with the specific discrepancy identified.
- 04
Extract from uploaded documents
Patients photograph insurance cards and referrals. An extraction pipeline reads them, validates against expected formats and flags low-confidence fields for staff confirmation rather than guessing.
- 05
Write back to the EHR
Completed intake writes into the record through the vendor's FHIR interface, eliminating transcription entirely. Anything the interface could not accept was surfaced as a short staff review queue rather than silently dropped.
The outcome
What changed afterward
The recovered front desk time was redirected to patient communication and scheduling, which the group reports as the more valuable outcome. Denial reduction produced a measurable revenue effect within the first quarter after rollout.
The document extraction pipeline has since been extended to referral processing and prior authorization support, using the same infrastructure and review workflow.
“We expected to save staff time. We did not expect the billing effect, which turned out to be the larger number.”
Keep exploring
Services used in this engagement
AI Automation
Automating the work that was too messy to automate before.
API Development
Interfaces other teams can build on without asking you questions.
UX Design
Research and structure, before anything gets styled.
More work
Other engagements
Logistics & Transportation20256 months
Replacing 400 status calls a day with a customer visibility portal
A national third-party logistics provider was losing bids over visibility capability and burning customer service capacity on status requests. We built a shipment tracking platform on top of their existing TMS.
- Inbound status calls within four months
- -62%Inbound status calls within four months
- Enterprise bids advanced past the visibility requirement
- 9 of 11Enterprise bids advanced past the visibility requirement
Ecommerce20254 months
A replatform that paid for itself in eleven weeks
A direct-to-consumer home goods brand had outgrown a heavily-app-laden storefront that loaded in six seconds on mobile. We rebuilt it, then worked the funnel.
- Mobile page load time
- 6.1s → 1.6sMobile page load time
- Mobile conversion rate
- +38%Mobile conversion rate
Next step
Have a problem shaped like this one?
Bring us the version of it you are dealing with. The first conversation costs nothing and usually clarifies more than a proposal would.
No pitch deck. A 30-minute conversation about what you are trying to achieve.