Designed and shipped a clinician-facing portal that consolidated five legacy systems behind a single, accessible interface — adopted by 92% of eligible clinicians in week one.
Anonymized delivery dashboard
Week-one adoption
92%
Task time
−41%
Systems consolidated
5 → 1
Before
Constraint
Build
Controlled cutover
After
Measured gain
Every engagement is framed around the business situation, the constraint that made it hard, and the decision that turned delivery into a controlled path to value.
Situation
Clinicians were toggling between five separate systems — EHR, lab, imaging, scheduling, and messaging — to complete a single patient encounter. Every previous attempt to unify the experience had failed on either adoption or compliance.
Constraint
The portal had to satisfy HIPAA, regional health authority requirements, and the accessibility needs of a workforce that ranged from digital natives to clinicians close to retirement.
Decision
Kept clinical systems of record in place and built a federated portal around the highest-volume tasks instead of forcing a risky data migration.
Build
We shadowed 22 clinicians across six sites and four specialties before drawing a single screen. The dominant insight: the design had to feel like an upgrade to the highest-volume task each role performed, not a unifier in the abstract.
Source-of-truth data stayed in the existing systems. The portal federated reads through a thin BFF and wrote back through audited adapters. No data was migrated — adoption did not depend on a cutover.
Outcome
92% of eligible clinicians used the portal at least daily in week one. 98% by week four.
Median time-on-task for a routine patient encounter dropped 41%.
What changed after launch
Clinicians started each shift from one accessible workspace, while compliance, audit, and PHI redaction moved into a single governed layer.
“The portal felt familiar on day one because the team designed around how clinicians actually move through an encounter.”
The work was not abstract modernization. It changed day-to-day behavior, ownership, and the evidence leaders used to make decisions.
Before
After
The delivery plan made the system boundary explicit, then used rehearsals, gates, and telemetry to optimize safely before launch.
Delivery architecture
Healthcare control loop
Discover
We shadowed 22 clinicians across six sites and four specialties before drawing a single screen. The dominant insight: the design had to feel like an upgrade to the highest-volume task each role performed, not a unifier in the abstract.
Build
Source-of-truth data stayed in the existing systems. The portal federated reads through a thin BFF and wrote back through audited adapters. No data was migrated — adoption did not depend on a cutover.
Launch
WCAG 2.2 AA conformance was a release gate, audited by an external partner before go-live. Keyboard-only flows for the top ten tasks were verified manually.
We shadowed 22 clinicians across six sites and four specialties before drawing a single screen. The dominant insight: the design had to feel like an upgrade to the highest-volume task each role performed, not a unifier in the abstract.
Source-of-truth data stayed in the existing systems. The portal federated reads through a thin BFF and wrote back through audited adapters. No data was migrated — adoption did not depend on a cutover.
The BFF gave us a clean place to enforce authorisation, audit, and PHI redaction in one layer.
WCAG 2.2 AA conformance was a release gate, audited by an external partner before go-live. Keyboard-only flows for the top ten tasks were verified manually.
After launch
Clinicians started each shift from one accessible workspace, while compliance, audit, and PHI redaction moved into a single governed layer.
Next.js
TypeScript
NestJS
PostgreSQL
Azure
Azure AD
OpenTelemetry
A similar problem?
A senior engineer will follow up within one business day with an opinionated take on the shape of the work.
