Emergency Deployment Recap Digitalization, Code for Canada × RCMP

Code for Canada x RCMP · Government

UX/UI Designer & Researcher

2022 – 2023

Code for CanadaRoyal Canadian Mounted Police

Digitizing the source of truth for RCMP deployments

Overview

The project focused on improving the internal workflows for RCMP officer scheduling and payments following emergency deployments. Initially, the problem presented as a massive backlog of error tickets in the scheduling system. Through discovery, it was revealed that the overtime (OT) submission portal lacked validation, allowing officers to submit incomplete or incorrect data.

The real breakthrough came when we zoomed out to map the entire end-to-end service — the root cause was an upstream, paper-based form called the Emergency Deployment Recap (EDR). The final solution pivoted to digitizing the EDR form within the secure intranet, with auto-population and validation to ensure data was clean before it ever reached the schedulers.

Key Results

  • Minimized error correction downstream
  • Reduced scheduling & payment ticket volume
  • Predictable, trackable digital workflow

Validation at the source eliminated manual correction downstream, cut the volume of scheduling and payment error tickets, and replaced a fragmented manual process with a trackable digital workflow.

Methods

  • User Research
  • User Journey Mapping
  • Competitive Analysis
  • Design Audit
  • Visual Design
  • Prototyping
  • Testing

Role & Timeline

Role

UX/UI Designer & Researcher

Timeline

October 2022 – October 2023 · 1 year

Where We Were Boxed In/ 03

Working inside a secure government intranet

Secure Environment Isolation

Required to work entirely within a restricted secure intranet with zero access to outside third-party technologies — no Google Drive, GitHub, or Figma-to-code tools.

Hardware & Access Latency

Significant delays in the onboarding process; discovery began well before secure laptops or physical keycards were issued.

Tooling Restrictions

Restricted to a narrow set of pre-approved applications; any new tool required a lengthy RCMP approval cycle that threatened the project timeline.

The Turning Signals/ 03

Early signals — a service failure, not a software bug

The Knowledge Gap

Identified a massive disparity in system knowledge between the scheduling team and the rest of the organization, leading to fragmented processes.

Critical Operational Failures

Payment volatility & scheduling latency

Officers were frequently experiencing extreme pay discrepancies — either significantly overpaid or not paid at all for overtime. Deployment schedules were consistently late, creating downstream stress for officers on the ground.

Entry Point Discovery

These early findings shifted the focus from a software bug to a service failure, proving that the human cost of the existing system was unsustainable.

The Symptoms/ 04

The initial challenge — operational fragility

The Crisis: Administrative Gridlock

The project was triggered by an enormous backlog of error tickets originating from the scheduling and overtime (OT) payment departments. Schedulers were effectively drowning in manual data correction.

The Initial Focus

The surface-level problem

The entry point for discovery was to improve the digital OT portal to enforce better data entry.

The Symptom

Schedulers were receiving a high volume of tickets because the internal OT submission portal lacked data validation.

The Impact

Officers could submit incomplete or incorrect entries, leading to immediate scheduling delays, payment delays, and a spike in support tickets.

Before state · Current submission process

Schedule Submission Flow

No input validation exists at the point of submission. Officers can send incomplete or error-filled schedules completely undetected — triggering a weeks-long correction and resubmission cycle before payment is ever processed.

OFFICERSCHEDULERYesNo2–6 WEEK DELAY PER CYCLEStartExtra Duty PayBRTOperationalScheduleOfficer inputsnew scheduleSubmit a Form⚠ NO VALIDATIONErrors pass through undetectedSchedulersreceiveCompletedproperly?Officerspaid ✓File rejected,resubmit

← scroll to explore →

Normal flow
Unvalidated handoff
Rejection / delay cycle
Success path

Initial research — the operational friction

Mapping the day-to-day surfaced a self-perpetuating loop. Each handoff fed the next: exhausted officers, a portal with no guardrails, and schedulers left to clean up after the fact — then back to the start, weeks later.

01Officer12+ hr post-deployment

The “Post-Shift” Reality

Comes off a high-stress emergency deployment — exhausted, with zero cognitive appetite for a complex laptop interface.

submits overtime
02OT Portalno validation layer

The Validation Gap

Nothing flags errors, so officers assume it’s handled and never review — incomplete, incorrect entries slip in unchecked.

bad data enters
03Schedulerfirst & only line of defense

The “Rejection Loop”

Catches the same repetitive errors, rejects the submission, and sends it back to the officer — to repeat all over again.

rejected — weeks later
↻ and the loop repeats

The Human Element

The Service Failure:“The Back-and-Forth”

The system wasn’t just slow — it was a revolving door of bad data.

4/5

Officers don’t understand the internal system.

100%

of officers interviewed had an OT payment rejected.

1000%

Frustration from both ends of the process.

Every interviewee used the word “HATE” at least once.

The friction & the pivot

Before moving upstream, we tested the obvious fix first — solving the problem right where the bad data entered the system.

Hypothesis A· Explored → set aside

A Google Chrome extension layered over the OT portal

The Goal

Stop bad data at the point of entry. Since the RCMP runs Google Chrome — and already permits a handful of approved extensions — a Chrome extension felt viable: it could add an immediate validation layer plus contextual, user-specific help on top of the existing OT portal, so exhausted officers got each submission right the first time without piling on cognitive load.

A Chrome extension layered over the BRT scheduling tool — a user-specific help panel on the left and a shift-code calculator on the right, guiding officers through OT entry without leaving the portal.

A Chrome extension layered over the BRT scheduling tool — a user-specific help panel on the left and a shift-code calculator on the right, guiding officers through OT entry without leaving the portal.

The Friction

Why we moved away

Getting a new Chrome extension approved meant clearing RCMP IT security — an approval cycle that would have outlasted our entire one-year project timeline. The idea was technically possible, but it simply couldn’t ship in time, so we had to pivot.

So we stopped, regrouped, and widened the lens.

Betting on the tools they already had

With a Chrome extension off the table, we looked at what officers could already reach inside their environment — and bet on AI assistance built on the RCMP's existing Microsoft stack.

Hypothesis B· Explored → de-prioritized

An LLM assistant layered onto the RCMP's Microsoft 365 environment

The Goal

Meet officers inside the tools they already used. Because the RCMP runs on Microsoft 365, Copilot was already present in their environment — so an LLM-assisted layer could walk officers through each submission and even auto-populate parts of the form from their field notes. Enabling it within the secure intranet promised a win well beyond this project: a trusted, reusable pattern for AI assistance across the RCMP's wider tech estate.

An in-context LLM assistant (Schedule Pal) layered onto the RCMP's existing scheduling interface — guiding officers through OT entry and surfacing contextual help without leaving the portal.

An in-context LLM assistant (Schedule Pal) layered onto the RCMP's existing scheduling interface — guiding officers through OT entry and surfacing contextual help without leaving the portal.

The Friction

Why we moved away

Partway in, we learned that a team in Eastern Canada was already building a closely related AI initiative. Rather than duplicate the work and create competing claims on the same stakeholders, we chose to de-prioritize our version — protecting the partnership and keeping our focus on the part of the problem only we were positioned to solve.

So we stopped optimizing the portal altogether and followed the data back to its source.

Shoulda, coulda, and woulda won’t get it done. In attacking adversity, only a positive attitude, alertness, and regrouping to basics can launch a comeback.

Pat Riley

The Turning Point/ 03

The “Upstream” Revelation

The Realization

Technical fixes to the portal were only treating the symptoms. Mapping the end-to-end flow revealed that the overwhelming majority of scheduling and payment errors were originating from a single paper-based form.

The Discovery

With further research — new insights and new information from user interviews — we discovered the Emergency Deployment Recap (EDR): a handwritten document filled out in a high-stress environment, where the data first lost its integrity.

The Pivot

We reframed the project from “portal optimization” to “digitizing the source,” moving the research and design work upstream to ensure data was clean before it ever touched the digital network.

The paper-based EDR form, filled out by hand during emergency deployments — forest fires, protests, large public events. Recreated here with a heavily refined aesthetic; the real form in use is roughly 10% as polished as this.

The paper-based EDR form, filled out by hand during emergency deployments — forest fires, protests, large public events. Recreated here with a heavily refined aesthetic; the real form in use is roughly 10% as polished as this.

Before state · Paper-based EDR process

Emergency Deployment Report (EDR) Flow

Officers returning from emergency deployments fill out a physical paper form using specialized hour codes — zero validation at the point of entry. Errors go completely undetected until a scheduler reviews the document days later, compounding into weeks-long payment delays with no digital trail to audit.

OFFICEREVENT SITE MANAGER / SCHEDULINGpaper deliveryYesNo2–4 WEEK DELAY PER REJECTIONOfficer Deployedto EmergencysituationDeploymentOverOfficers requestEDR form⚠ PAPER — ZERO VALIDATIONWrite hours inspecialized code+ type of hourHand-form toget approvedSubmit to EventSite ManagerWait for it tobe verifiedVerdict?Approved ✓Submit Confirmedform to SchedulingDepartmentOfficerspaid ✓Rejected

← scroll to explore →

Normal flow
Paper form / unvalidated handoff
Rejection / delay cycle
Success path

90%

of tickets originated from the paper EDR form.

100%

paper-based process, end to end.

0

data validation steps before digitization.

Phase 2 Discovery

The root cause — the EDR form

Moving upstream to solve the real problem.

The Revelation

These errors were never caused by the digital portal itself. The problem existed upstream — at the data source.

Who's in the loop

Officers

8–12 hour shifts on a 4-on / 4-off rotation, plus unpredictable emergency deployment shifts.

Supervisors

On the ground at emergency sites — signing off on forms and overseeing officers.

Schedulers

The first and only line of defense against errors, with no automation or validation to back them up.

End-to-end workflow analysis

A single bad form didn't stay contained — every handoff downstream amplified the damage.

01

Paper EDR Form

Root Problem
  • Paper-based, handwritten forms
  • High-stress field environment
  • Incomplete & illegible data
  • Delayed submission
02

Manual Entry

Error Amplification
  • Bad data transcribed into the system
  • Interpretation errors
  • Missing information
03

Scheduling Team

Headaches & Backlog
  • Physical boxes of “complete” EDR forms
  • Error tickets pile up
  • No time to catch up
Error volume
Amplified ↑

Old Approach

Patch the internal portal to catch & fix errors downstream.

New Approach

Digitize the EDR form to ensure data integrity at the point of origin.

The Transformation

From paper to a validated digital recap

The same recap, reimagined at the source — the handwritten EDR form replaced by a structured, validated digital submission.

Before — The paper EDR

Before/ The paper EDR

  • Handwritten and frequently illegible
  • Incomplete or skipped fields, no checks
  • Filled out in high-stress field conditions
  • Weeks-long back-and-forth rejection loop
  • Re-keyed by hand — transcription errors downstream
After — The digital EDR

After/ The digital EDR

  • Auto-populated from officer & deployment records
  • Hard validation blocks bad data at entry
  • Legible, structured, consistent every time
  • One-click approval flow for schedulers
  • Zero manual correction needed downstream
After state · Schedule Pal digitized solution

Schedule Pal User Flow

The digitized solution eliminates every failure point of the paper EDR. Fields are auto-populated, hours are auto-converted to code, and a validation test catches errors before submission — collapsing a weeks-long delay cycle into minutes.

✦ Auto-populated fields✦ Automatic code conversion✦ Validates before submission✦ Errors fixed in minutes, not weeks
YesNoNoFix Errorscaught in-app — resolved in minutes, not weeksOfficer Login toRCMP CertifiedComputerClick SchedulePal linkInfo already logged(ID, Email, Name)✦ auto-populatedInput hoursAuto-converts to code+ Site Cmdr Email✦ no manual codes neededCorrectlyfilled out?ValidationTest✦ before submissionAuto EmailsSite Commanderto approveAuto-submitsto ScheduleDept ✓

← scroll to explore →

Standard flow
Smart / auto step
Validated success path
Fix Errors (in-app, caught instantly)

The final solution — the digital EDR

Digitizing the Source of Truth

Instead of patching the internal portal, we built a custom digital form directly within the secure RCMP intranet.

Upstream Logic

By moving the design work to the point of origin — the EDR — we ensured that only clean data could enter the scheduling system.

Modular, role-aware dashboard

Modular, role-aware dashboard

A modular dashboard built for ease of use and easy to extend as new features land. Because officers are already signed into the secure intranet, it recognizes whether you're an officer, site commander, or scheduler — and tailors the view to each role.

Recap submission form

Recap submission form

Auto-scrapes intranet records to pre-fill personal details, with a modular, responsive layout. Keeping the whole recap on one page lets officers review missed dates, fix errors, add recipients, and send it off — while a history log tracks each form from created to in progress, rejected, or approved.

Admin page

Admin page

Schedulers see every form awaiting attention in one queue. Once a recap is approved and pushed to the internal portal, the completed form can be cleared from the list.

Key design decisions — solving for friction

01

Reducing cognitive overhead

Auto-Population & Field Mapping

Leveraged existing intranet data to auto-fill officer details, reducing repetitive typing for officers coming off high-stress, 12-hour shifts.

02

Blocking bad data at entry

Hard Validation Rules

Implemented strict formatting and required-field logic to block incomplete or illegible submissions before they could reach a scheduler.

03

Structuring for exhausted users

Cognitive Load Reduction

Structured the form into clear, logical sections so exhausted users could navigate quickly without errors.

04

Ending the rejection loop

One-Action Approval Flow

Designed a streamlined view for supervisors to approve or deny with a single click, giving immediate feedback to the officer and ending the weeks-long rejection loop.

The outcome

100%

clean data

Data Integrity

Replaced a manual, error-prone paper process with a consistent, trackable digital workflow.

0

Manual Corrections

System Efficiency

Ensured data reached schedulers in a format that required zero manual correction, effectively killing the support ticket at the source.

100%

time saved

Process Improvement

Eliminated the weeks-long rejection loop, providing immediate feedback and dramatically reducing turnaround time.

Shipping Without the Usual Tools/ 03

Technical constraints & handoff

Back to the Stone Age

The constraints

No shared repositories, no Figma-to-code plugins, and no cloud-based collaboration tools due to the secure intranet.

The Process

Conducted line-by-line synchronous code reviews with the developer to preserve design logic, and manually wrote CSS specifications and asset handoffs to match the original design intent without modern automation tools.

The Takeaway

Proved that design integrity can be maintained through direct, disciplined communication and technical pragmatism.

How We Measured It

Measurable outcomes

Method 01

Live Deployment Monitoring

Manual ticket review + portal analytics

Outcomes were validated during a real, active deployment — a forest fire. We manually reviewed every ticket that came through during that period and cross-referenced it with the internal portal's built-in analytics tools. This gave us a ground-truth picture of what the system was doing under real conditions, not test conditions.

Method 02

Post-Event User Interviews

Officers & Schedulers — both sides of the loop

After the deployment wrapped, we interviewed both user groups to surface any friction and make rapid fixes. Both sides gave strongly positive feedback about the digitization itself. The recurring theme: even officers who were skeptical of switching from paper said the digital form was significantly less stressful under pressure.

The Biggest Outcome

Full elimination of the paper EDR form.

No hybrid, no parallel process — the digital form became the only path. More significantly, this project influenced the RCMP to initiate a broader digitization push across their other paper-dependent forms, using our work as the blueprint for reducing friction organization-wide.

Project Debrief

2

Hypotheses explored and set aside — each one sharpened the real problem

1

Paper form responsible for an entire department’s operational breakdown

100%

Of the solution came from reframing the brief and following the data upstream

The best design solution for a complex system isn’t always a new feature; sometimes it’s moving the work three steps upstream.

Project Reflection