A platform people worked around
R1 Decision is the core workflow tool for thousands of associates, supervisors and managers in R1 RCM’s India operations, used for denial management and accounts-receivable follow-up — a process covering over 400 million tasks a year. In mid-2024, senior leadership asked for one thing: increase adoption of the platform’s built-in tools.
My role: create the conditions
As Product Design Manager leading the India Design Organisation, I ran the project — a design team, stakeholders across product, engineering and operations, and six weeks to deliver on a commitment leadership had already made. My job wasn’t to design the solution myself. It was to create the conditions for the right solution to emerge, and make sure the team had what it needed to deliver it.
How I ran it
- Turn the mandate into a briefRan a kickoff with product, engineering and operations to agree measurable goals — less tool-switching, less time on task, fewer errors — and name constraints and unknowns.
- Go to the floorPlanned contextual inquiry, interviews, a heuristic evaluation and a journey-map audit, and joined sessions in Noida and Gurgaon myself.
- Synthesise togetherFacilitated synthesis with product and operations so the direction belonged to everyone; each opportunity got testable acceptance criteria.
- Handle the constraint in the openWhen engineering confirmed the APIs couldn’t support real-time updates, I used an impact-vs-complexity matrix to make the scope call with stakeholders.
- Direct the workConcept workshop, then wireframes through hi-fi prototypes to usability testing with associates, supervisors and managers.
- Ship it rightHandoff, specs and QA ran through me, so what was built matched the designs and the reasoning behind them.
What research showed
Users weren’t avoiding features because they disliked them. They avoided them because the architecture made workarounds faster. Associates kept seven tabs open because no single screen gave them what they needed to decide. Notes went into Word documents, and high-value accounts lived in personal spreadsheets.
If auth status, clinical notes and denial reason were all in R1, I could use fewer tabs.Associate, Noida
The heuristic evaluation backed this up: steps lacked context, financial data was buried in dense tables, consequential actions had no validation, and the call queue gave no visibility into status.
The solution: one screen to decide
A three-panel interface. Patient identity, claim status, financial data, eligibility and a fully integrated notes history sit on one screen. Notes became contextual and auto-saved within the workflow, and status badges in the call queue mean handovers no longer require rebuilding context.
Decisions and trade-offs
Near-real-time instead of real-time
Without real-time APIs, structured prompts replaced live flows — keeping the user value without backend changes we couldn’t get. Full automation moved to a later quarter.
Notes as the quiet centrepiece
Not the most visible problem, but fixing it changed behaviour fastest, so it got priority.
Pre-grooming with engineering
A concept-stage review before formal grooming surfaced technical concerns early. Both teams adopted it as standard practice.
What I learned
Framing the problem before the team touched a design tool was the highest-leverage thing I did. “Increase feature adoption” and “fix the architecture” lead to different solutions — and only one changes behaviour.