contact central. 360 desktop - advocate experience.

client health insurance payer (anonymised)
industry healthcare · health insurance
duration 2025 · six months

my role lead ux architect

platform pega customer service · constellation
tools figma, figjam, miro
methods stakeholder workshops, data analysis, ivr workflow mapping, personas & jtbd, journey mapping, service blueprint
Contact Central 360 advocate desktop shown in the dark theme
Contact Central 360 advocate desktop shown in the light theme
light dark

drag to compare the dark & light advocate desktop, both ship wcag 2.1 aa

see it live

project overview

a health insurance payer's advocates resolved complex member enquiries across a wall of disconnected systems: eligibility, claims, benefits, records, knowledge, and correspondence each in a separate tool. every call meant repeated logins, searches, and verification, and the operating model made speed and accuracy compete.

in healthcare that trade-off is expensive. a missed grievance signal or a wrong waiting-period answer is a compliance and member-harm risk, not a rounding error, yet advocates were expected to "just know" dense, changing policy with guidance detached from the moment of need.

i led the discovery and definition for contact central 360: one context-aware advocate desktop on pega customer service and constellation, holding a single member context from first question to fulfilment. i facilitated the workshops, ran the data and journey analysis, mapped the ivr and call flows, and defined the personas and design decisions.

one interaction, one context. the member experiences a single conversation, so the advocate should work in one too.
research& findings

research & findings

discovery workshop

discovery began with the people who live the problem. i facilitated a cross-functional workshop with advocates, supervisors, operations, compliance, knowledge, and pega architecture, using the members' real call drivers as the agenda rather than a feature wishlist. together we walked the current call end to end, marked every failure point and workaround, and prioritised them by member harm, compliance risk, volume, and effort. the outputs became the backbone of the whole engagement.

Discovery workshop synthesis — cross-functional participants (advocates, supervisors, operations, compliance and privacy, knowledge, pega architecture, product, design); a four-step process (align on outcomes and scope, walk the current call journey, mark failure points and workarounds, prioritise by member harm, compliance, volume and effort); and outputs: validated personas and JTBD, current-state journey and service blueprint, prioritised pain-point backlog, IVR and call-workflow maps, and first-cut Pega case types and integrations

research & findings

the people we designed for

three primary personas shaped every decision: the advocate who resolves the call, the member the call is about, and the caller on the line, who may be the member, an authorised representative, a caregiver, or a provider. because authority and relationship vary, identity and authorisation come first, and the conversation adapts to who is calling.

Three primary personas — the Advocate (primary user: contact-centre agent handling inbound, multi-system calls under handle-time pressure), the Member (health-plan member with varied literacy and accessibility needs, often stressed), and the Caller/Contact (member, authorised rep, caregiver or provider, verification-first) — each with context, goals and needs

research & findings

jobs, pains, and gains

for each persona we mapped the jobs to be done, the pains that slow them down or create risk, and the gains worth designing for. one rule cut across all three: identity and authorisation are established before anything else.

Jobs, pains and gains matrix across the three personas — Advocate (primary user), Member (member/patient), and Caller/Contact (rep). Jobs to be done, pain points, and gains for each, with a footer note that identity and authorisation are established before anything else

research & findings

making sense of member data

every call opens with the same problem: find the right person fast, then read their record correctly. i analysed how advocates search, what the member record holds, and how enrolment and eligibility drive what an advocate can answer, then defined the order the desktop should reveal it.

research & findings

how a call finds its way

long before it reaches an advocate, the ivr shapes a call: what the member hears, how they are identified, what they can settle on their own, and how everyone else is routed. i mapped the call flows end to end so the desktop understands the context a call already carries the moment it lands.

main ivr menu

greeting, language, and the intents a caller can pick from

Main IVR menu flow — greeting and language choice, then the top-level intents (claims, benefits and eligibility, ID card, authorisation, billing, something else), each routing to self-service or to an advocate queue, with repeat and accessibility options

caller authentication

verifying identity before anything sensitive is shared

Caller authentication flow — capturing member or subscriber ID and verifying identity by date of birth and security details, with matched, partial-match and failed paths, a knowledge-based fallback, and the verified identity handed to the advocate desktop

self-service & deflection

resolving common tasks automatically, with a clean hand-off when needed

Self-service and deflection flow — automated resolution for common intents such as claim status, eligibility check, ID card request and address change, with success confirmation and fall-through to an advocate when the caller opts out or the task exceeds self-service

route to advocate

skills-based routing that lands the call on the right advocate

Route-to-advocate flow — skills-based routing by intent and plan, priority and queue selection, wait-time and callback offer, and the screen-pop that hands the verified member and reason-for-call to the advocate

callback & after-hours

queued callbacks and after-hours handling, so no call is dropped

Callback and after-hours flow — queued callback capture and scheduling, and after-hours handling for emergencies, self-service-only intents, message capture and next-business-day callback, so no call is simply dropped
ideation& designthinking

ideation & design thinking

the call, in the member's shoes

to design the desktop i first mapped the call as the member and advocate actually live it: the trigger, the ivr, the wait, identity checks, the back-and-forth across systems, and the resolution, or the callback that never comes. i charted phases, actions, thoughts, and emotion, and marked the moments that make or break trust, so every later design decision could point back to a real point of pain.

Current-state customer journey map — six phases (trigger and need, ivr and routing, authenticate, explain the problem, advocate works across systems, resolution and follow-up) with rows for member actions, touchpoints, thoughts and an emotion curve that dips lowest at authentication and while the advocate switches between systems; pain-point chips (repeated identity, long holds, tool switching, uncertain answers, dropped callbacks) and opportunity chips (single member context, guided verification, answers with source, reliable callback)

ideation & design thinking

the choreography behind the call

a journey shows what the member feels; a blueprint shows what it takes to deliver. i mapped the layers behind a single call: frontstage advocate actions, backstage steps, the systems each one touches, and the policies and hand-offs in between. the lines of interaction and visibility made the hidden effort obvious, and pointed straight at what the desktop had to unify.

Current-state service blueprint — the same six call phases mapped across lanes for physical evidence, member actions, a line of interaction, frontstage advocate actions, a line of visibility, backstage actions, support systems (pega customer service, eligibility, claims, benefits, member records, knowledge base, correspondence) and policies and hand-offs, with coral markers flagging the friction points where an advocate must swivel between disconnected systems

ideation & design thinking

user stories & use cases

the research turned into five advocate use cases, each framed as a user story and a short flow. together they set the scope for the desktop: what an advocate has to accomplish on a call, and the safeguards the interface has to enforce.

identity verification

verify an ivr-routed caller before any case data is exposed.

as an advocate, i want the case to pop and identity pre-checked so i can help a verified member without re-asking basic info.

the flow at a glance
1screen pop on ani match, ivr context carried in
2verify 2 of 3, ivr already matched date of birth
3case unlocks, verified-by and time stamped
sensitive panes masked launch actions locked reject → limited-info mode
lo-fi wireframe

eligibility & benefits

answer eligibility and benefits questions with accurate, current plan data.

as an advocate, i want a benefits view pre-filled from the open case so i can quote coverage, copays and accumulators without switching systems.

the flow at a glance
1edb drawer opens, member pre-filled from the case
2quote coverage, copays and accumulators in one view
3knowledge buddy ranked by the ivr intent
single source of truth no system switching
lo-fi wireframe

claim status & dispute

explain claim status and denials, and start a correction when a member disputes one.

as an advocate, i want claim and eob detail plus a guided reprocess path so i can resolve disputes in one call instead of a callback.

the flow at a glance
1claims filtered to the call reason, denial surfaced first
2read claim and eob detail together
3guided reprocess with routing, sla and reference number
resolved in one call reference number for the member
lo-fi wireframe

prior authorization

submit and track a prior authorization for a service that requires one.

as an advocate, i want auth intake to inherit the member and benefit data already on the case so i can file a compliant request without leaving the call.

the flow at a glance
1launch auth as a sub-tab, call controls stay live
2criteria checklist, case data inherited
3outcome: reference number, due date, member read-back
call stays live data inherited, no re-keying
lo-fi wireframe

grievance & appeals

file a grievance or appeal within compliance windows and evidence requirements.

as an advocate, i want filing-deadline and evidence checks built into intake so i never submit a case that compliance would reject later.

the flow at a glance
1intake pre-linked to the denied claim, filing-window guard rail visible
2submission gate: nothing files until the compliance checklist is satisfied
3confirmation: sla clock, owner queue, next member touchpoints
filing-window guard rail compliance gate before submit
lo-fi wireframe
challenge& achievements

challenge & achievements

the operating model made speed and accuracy compete, inside a live telephony platform and under strict compliance. the desktop had to make them reinforce each other instead, without inventing a new system for advocates to learn.

challenge

a wall of disconnected systems

advocates moved between eligibility, claims, benefits, records, knowledge and correspondence one tool at a time, re-searching and re-verifying every call.

response

one member context

the open case carries the member across every task, so the advocate reads one record instead of assembling it from six.

challenge

speed versus accuracy

the model rewarded handle time, but a wrong benefit or waiting-period answer is a member-harm and compliance risk, not a rounding error.

response

guidance at the moment of need

guided flows surface the right data and ranked knowledge in place, so getting it right stops costing time.

challenge

rules held in people's heads

advocates were expected to just know dense, changing policy, and to remember verification and filing rules under pressure.

response

safeguards built into the interface

the verification gate masks data until identity is confirmed and the compliance gate blocks a filing until its checklist and window are satisfied. the interface enforces the rules, not memory.

challenge

a modern experience inside pega constellation

the desktop had to feel calm and modern while living within the client's design system and platform constraints.

response

patterns the client can own

built within existing components and case types, shaping reusable patterns that scale beyond these five flows.

the result: a single advocate desktop that unifies the member context, embeds the compliance safeguards, puts guidance at the point of decision, and extends the client's design system rather than replacing it.

challenge & achievements

key interactions, up close

identity verification gate

the unlock: no member data or high-risk action until the caller is verified.

annotated interaction

compliance submission gate

the guard rail: nothing files until the deadline and evidence checks pass.

annotated interaction
overview& takeaway

overview & take away

Storyboard summarising the Contact Central 360 project: disconnected systems as the problem, the research that set the scope, the five advocate use cases, the verification and compliance safeguards, the unified single-context desktop, and the calmer, compliant call as the outcome

overview & take away

the through-line

cc-360 turned a fragmented, high-risk advocate workflow into one guided, compliant desktop. research set the scope, ideation shaped the flows, and the interface carried the safeguards, so speed and accuracy stopped competing.

what this project taught me

in a compliance-heavy domain the strongest design moves are the ones that take the rules out of people's heads and put them into the interface. the verification gate and the compliance gate are not features, they are the product doing the remembering so the advocate can stay with the member.

the desktop's job was not to show more. it was to protect the member and the advocate at the two moments that matter most, and to make the safe path the easy one.

the measure that mattered was never a screen count. it was whether an advocate could stay present with a worried member while the desktop quietly handled identity, policy, and compliance in the background. get that right and the technology disappears, which in a contact centre is the highest praise the work can earn.