arrow_back Back to Projects
Case Study Program Delivery · Quality

Interactive Data Dashboard

Ran a large, multi-workstream healthcare integration program and built the live view that told stakeholders whether delivery was on track, whether issues were caught early, and what was driving them. One place to see where every workstream stood and what to fix next, instead of days-long manual status and defect reports.

RoleTechnical Program Manager
StackSmartsheet · Notion
AudienceEng, QA & ops stakeholders
OutcomeOn-track delivery, real-time reporting
info

Figures tagged Illustrative are representative examples that recreate the concept, not real data.

The Brief

The business questions it answered

Leadership didn't need more data. They needed answers. I built the dashboard around the specific recurring questions stakeholders kept asking about delivery and quality:

help

"Are we keeping up? Are we closing bugs faster than we open them?"

help

"Are we catching issues early in review, or late from customers?"

help

"What are the top root-cause themes, and are they trending up?"

help

"How long are defects taking to close?"

help

"Where does every workstream stand, and what is blocked right now?"

Running Delivery

How the program moved

Behind the metrics was the board that ran the work. Every integration workstream moved through one delivery lifecycle, from discovery through to live, each with an owner and a target. Blockers were not a stage. They were flagged on whatever stage a workstream sat in, so problems surfaced early and nothing stalled in silence.

view_kanbanDelivery board
Illustrative
Discovery chevron_right Design chevron_right Build chevron_right QA chevron_right UAT chevron_right Live

~50

In flight

9

In QA

4

Blocked

Provider portal order entryLiveIntegrationsQ1
Results delivery serviceUATEngQ2
Allergy & medication moduleQAQAQ2
Custom panel mappingblockBlockedBuildIntegrationsQ2
Scheduling syncBuildEngQ3
Two-way notes syncBuildEngQ3
Access & two-factor sign-inDesignEngQ4
Document & attachment handlingDiscoveryIntegrationsQ4
flag

Program roadmap

check_circle

Core order & result flow

Inbound orders and outbound results live over the standard interface.

arrow_forward
schedule

Scheduling & notes

Appointment sync and two-way clinical notes.

arrow_forward
schedule

Self-serve provisioning

Partners onboard new connections without manual setup.

arrow_forward
schedule

Modern standards & analytics

Move toward modern interoperability and richer reporting.

Program completion~60%
A recreated, anonymized view of the delivery board and roadmap. Workstreams, owners, and dates are illustrative, with no real company, vendor, or personnel data.
warningRisks & dependencies
Illustrative
Upstream interface change on a partner connectionOpenIntegrations
Shared environment contention during release windowsMitigatedEng
Third-party mapping turnaround on new panelsOpenIntegrations
A recreated, anonymized risk log. Items, statuses, and owners are illustrative, with no real company, vendor, or personnel data.
How I Ran It

The operating cadence

The board was only as good as the rhythm behind it. As the program manager, I owned intake, kept the cross-functional view honest, drove blockers and dependencies to resolution, and reported on a steady cadence so stakeholders always knew where things stood.

inbox

Intake & prioritization

Sized incoming work against capacity and ranked it against program goals.

groups

Cross-functional triage

Brought engineering, quality, and operations to one view of status and severity.

link

Dependency & blocker management

Tracked risks and dependencies, and drove blockers to an owner and a date.

campaign

Reporting cadence

Published a regular, self-serve status so stakeholders did not wait on a manual report.

Triage

From intake to insight

Every defect landed here first, tagged by where it came from and how urgent it was. This is the raw intake the dashboard below turned into trends: triaged on this board, measured in that one.

bug_reportDefect triage
HighMedLow
Illustrative

To triage

5

Status not refreshing in portal after update

Portal

Report not generated on schedule

Support

In progress

3

Mapping mismatch on result import

LIS

Bulk download intermittently fails

Portal

In review

2

Retry storm on upstream timeout

LISblockBlocked

Notification not sent on later orders

Support
A recreated, anonymized triage board. Defect titles and tags are illustrative, with no real patient, company, vendor, or personnel data.
The Artifact

The dashboard

monitoringQuality & Delivery
Live Illustrative

Open defects

128

across all areas

Net this week

-6

closed > created

Early-detection

68%

caught in review

Median close

4.2d

time to resolve

Inflow vs outflow

trending_up

Are we keeping up? Created vs closed per week.

CreatedClosed8 wk

Detection: early vs late

verified

Catching issues before customers do.

Caught in internal review68%
Customer-reported32%

trending_upEarly share up from 54% over the quarter.

Top root-cause themes

category
Mapping / configuration34%
Data validation26%
Timeout / retry21%
UI state19%
A recreated, anonymized view of the dashboard concept. Numbers are illustrative, with no real company, vendor, or personnel data.
What It Tracked

The metrics on the board

Illustrative

Representative metric set, recreated to show the approach.

sync_alt

Created vs closed

Net defect flow / week

verified

Early-detection rate

% caught before customers

category

Root-cause mix

Top defect themes

timeline

Time to close

Median resolution time

Impact

How it changed decision-making

Before

Status and defect trends were assembled by hand and reviewed late. Blockers and problems were often noticed only after they had grown, and by the time a report landed, the moment to act had passed.

After

The team self-serves one live view of where delivery stands and what is breaking. Conversations shifted from "what is the status?" and "what broke?" to "are we on track, trending safe, and what do we fix next?"

Illustrative

Representative outcomes, recreated to show the shift.

bolt

Reporting turnaround

~[2–3] days to real-time

verified

Early-detection rate

54% to 68%

event_available

On-time delivery

~[90]% on or ahead of target

Reflection

What I would carry forward

insights

Instrument earlier

The reporting view paid off most when it existed before the work scaled, not after. I would stand it up at the start next time.

visibility

Make blockers visible by default

Treating blocked as a flag on the real stage, not a stage of its own, kept the lifecycle honest and surfaced risk without hiding progress.

self_improvement

Let the team self-serve status

Once status was live and shared, my time moved from assembling reports to removing obstacles, which is where a program manager adds the most.

arrow_backAll projects mailDiscuss this work