Narrative SOP — Pyze Execution Intelligence

Regulatory Context Narrative content is the primary artifact regulators read when evaluating a case. Governed by ICH E2B(R3) and subject to strict compliance review.

Why this SOP is audit-grade Every step below is captured at the DOM element level (button labels, field IDs, section panels) — not as screenshot coordinates. Every metric is derived from 2,211 real events across 550 cases, with confidence and case evidence visible per step.

Step Journey

Click a task to see the application journey on the same case, then click an application to drill into the execution detail Pyze captured for it.

Level 2 · Application Journey

Open Narrative Workspace — 10 apps touched within ±15 min of this task on the same case

Case counts per app are not mutually exclusive — a case that touched both Veeva and Phobos in this window is counted in both rows. Overlap reflects real cross-app work, not double-counted cases.

Level 3 · Execution Detail

Veeva Safety

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <span> "Narrative" 860×

pyzeClick <span> "Complete" 835×

pyzeClick <a> "Show more" 603×

pyzeClick <span> "Adverse Events" 388×

pyzeClick <span> "Products" 235×

pyzeClick <span> "Documents" 205×

pyzeClick <span> "Submissions & Distributions" 205×

pyzeClick <button> "Save" 188×

Level 3 · Execution Detail

Phobos

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <button> "Proceed" 137×

pyzeClick <button> "Save Comments" 93×

pyzeClick <button> "Advance" 82×

pyzeClick <button> "Send to process" 69×

pyzeClick <select> "Select Compound · - sitagliptin phosphate Tablet (G" 49×

pyzeClick <a> "Cases" 45×

pyzeClick <select> "Select Attachment description · Legal · Legal Pre-Sca" 38×

pyzeClick <a> "Cases Coordinator" 34×

Level 3 · Execution Detail

Microsoft Word

Document-level capture

Task Mining captures the specific file open during work — file names often encode case context.

file V-Narrative 543823 (English) (26544362_0_2_343646) 873×

file V-Narrative 533164 (English) (26378749_0_4_342805) 659×

file V-Narrative 533533 (English) (26384397_0_2_337752) 361×

file V-Narrative 551229 (English) (26664901_0_1_349465) 326×

file V-Narrative 549707 (English) (26645831_0_1_347236) 324×

file V-Narrative 552979 (English) (26689105_0_2_348832) 307×

file V-Narrative 542512 (English) (26522671_0_2_344761) 255×

file V-Narrative 530932 (English) (26349278_0_2_339374) 248×

Level 3 · Execution Detail

Acrobat

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

3,090 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 3 · Execution Detail

Microsoft Teams

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

2,449 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 2 · Application Journey

Select Narrative Template — 10 apps touched within ±15 min of this task on the same case

Level 3 · Execution Detail

Veeva Safety

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <li> "Narrative" 683×

pyzeClick <span> "Complete" 593×

pyzeClick <a> "Show more" 520×

pyzeClick <li> "Details" 288×

pyzeClick <button> "Create" 131×

pyzeClick <li> "Patient" 122×

pyzeClick <span> "Validation Results (Failures & Warnings)" 117×

pyzeClick <span> "Medical History & Concurrent Conditions" 115×

Level 3 · Execution Detail

Phobos

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <button> "Proceed" 153×

pyzeClick <button> "Advance" 89×

pyzeClick <button> "Save Comments" 77×

pyzeClick <div> "Please reply to “Request Information" using dropd" 52×

pyzeClick <button> "Send to process" 43×

pyzeClick <select> "Select Compound · - sitagliptin phosphate Tablet (G" 40×

pyzeClick <a> "Cases" 35×

pyzeClick <select> "Select Attachment description · Legal · Legal Pre-Sca" 34×

Level 3 · Execution Detail

Microsoft Word

Document-level capture

Task Mining captures the specific file open during work — file names often encode case context.

file V-Narrative 540591 (English) (26495494_0_2_345212) 726×

file V-Narrative 533164 (English) (26378749_0_4_342805) 659×

file V-Narrative 534252 (English) (26399499_0_2_343143) 572×

file V-Narrative 553016 (English) (26689302_0_2_348778) 392×

file V-Narrative 530932 (English) (26349278_0_2_339374) 387×

file V-Narrative 528191 (English) (26300681_0_1_338349) 382×

file V-Narrative 533533 (English) (26384397_0_2_337752) 361×

file V-Narrative 525025 (English) (26253159_0_3_342205) 320×

Level 3 · Execution Detail

Acrobat

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

2,900 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 3 · Execution Detail

Microsoft Teams

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

2,339 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 2 · Application Journey

Open Narrative Document — 10 apps touched within ±15 min of this task on the same case

Level 3 · Execution Detail

Veeva Safety

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <span> "Complete" 474×

pyzeClick <span> "Narrative" 319×

pyzeClick <a> "Show more" 160×

pyzeClick <span> "Adverse Events" 149×

pyzeClick <span> "Details" 135×

pyzeClick <span> "Validation Results (Failures & Warnings)" 134×

pyzeClick <button> "Save" 133×

pyzeClick <li> "Narrative" 132×

Level 3 · Execution Detail

Phobos

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <button> "Proceed" 72×

pyzeClick <button> "Advance" 63×

pyzeClick <button> "Save Comments" 43×

pyzeClick <div> "Please reply to “Request Information" using dropd" 43×

pyzeClick <select> "Select Compound · - sitagliptin phosphate Tablet (G" 36×

pyzeClick <select> "Select Report Type · Legal · Literature · Medical Recor" 24×

pyzeClick <button> "Send to process" 16×

inputld <select> "Select Compound · - sitagliptin phosphate Tablet (G" 16×

Level 3 · Execution Detail

Microsoft Word

Document-level capture

Task Mining captures the specific file open during work — file names often encode case context.

file Narrative 526208 (English) 192×

file V-Narrative 531868 (English) (26362311_0_1_340534) 185×

file Narrative 479307 (English) 184×

file Narrative 534695 (English) 169×

file Narrative 523385 (English) (1) 161×

file V-Narrative 531284 (English) (26355806_0_2_340584) 156×

file V-Narrative 531892 (English) (26362408_0_2_340575) 156×

file V-Narrative 531929 (English) (26362464_0_2_340629) 154×

Level 3 · Execution Detail

Microsoft Teams

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

993 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 3 · Execution Detail

Microsoft Outlook

Message-level capture

Task Mining captures email subjects. Case numbers frequently appear in the subject line.

subj <Due today><Void request><GPVCM POLAND> Case # 2326062 23×

subj <GPVCM Announcement> Promotions to Safety Data Manager positions in Poland 18×

subj Due today reports on 03-DEC-2025 14×

subj Due today reports on 01-DEC-2025 13×

subj Untitled 8×

subj [Confidential]<GPVCM Poland> <Completed>Escalation is Completed; Title: 2352930; Area: Post Marketed; ID: 4890 6×

subj Due today reports on 27-NOV-2025 4×

subj <Due today><Void request><GPVCM POLAND> Case # 4×

Level 2 · Application Journey

Generate Narrative (AI-assisted) — 10 apps touched within ±15 min of this task on the same case

Level 3 · Execution Detail

Veeva Safety

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <span> "Complete" 145×

pyzeClick <a> "Show more" 75×

pyzeClick <span> "Generate Narrative from Outline" 65×

pyzeClick <li> "Narrative" 60×

pyzeClick <span> "Narrative" 48×

pyzeClick <button> "Create" 40×

pyzeClick <span> "Submissions & Distributions" 34×

pyzeClick <li> "Unknown" 33×

Level 3 · Execution Detail

Phobos

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <button> "Proceed" 21×

pyzeClick <button> "Advance" 15×

pyzeClick <button> "Save Comments" 13×

pyzeClick <select> "Select Compound · - sitagliptin phosphate Tablet (G" 9×

pyzeClick <button> "Send to process" 6×

pyzeClick <a> "Cases" 5×

inputld <select> "Select Compound · - sitagliptin phosphate Tablet (G" 5×

pyzeClick <button> "Send to Medical Review" 4×

Level 3 · Execution Detail

Microsoft Word

Document-level capture

Task Mining captures the specific file open during work — file names often encode case context.

file V-Narrative 525025 (English) (26253159_0_3_342205) 211×

file V-Narrative 542512 (English) (26522671_0_2_344761) 189×

file V-Narrative 531284 (English) (26355806_0_2_340584) 156×

file V-Narrative 523170 (English) (26221379_0_2_340465) 144×

file V-Narrative 433845 (English) (24804813_0_2_345392) 137×

file V-Narrative 530932 (English) (26349278_0_2_339374) 118×

file V-Narrative 537919 (English) (26452534_0_2_341686) 99×

file V-Narrative 498632 (English) (25800349_0_2_340329) 87×

Level 3 · Execution Detail

Microsoft Teams

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

534 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 3 · Execution Detail

Acrobat

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

390 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 2 · Application Journey

Review Narrative — 10 apps touched within ±15 min of this task on the same case

Level 3 · Execution Detail

Veeva Safety

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <span> "Complete" 99×

pyzeClick <div> "Narrative" 65×

pyzeClick <a> "Show more" 44×

pyzeClick <div> "Case Assessment Results" 44×

pyzeClick <span> "Adverse Events" 35×

pyzeClick <button> "Save" 33×

pyzeClick <div> "Case Assessments" 33×

pyzeClick <span> "Products" 32×

Level 3 · Execution Detail

Phobos

DOM-level capture

Pyze JS captures every element interaction — button labels, field IDs, section panels.

pyzeClick <button> "Proceed" 21×

pyzeClick <button> "Send to process" 19×

pyzeClick <button> "Advance" 14×

pyzeClick <td> "2345469 (v5.1) Japanese (Japan)" 12×

pyzeClick <button> "Save Comments" 10×

pyzeClick <td> "2352671 (v0.1) Japanese (Japan)" 10×

pyzeClick <button> "Login" 9×

pyzeClick <td> "2347576 (v2.1) Japanese (Japan)" 6×

Level 3 · Execution Detail

Acrobat

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

349 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Level 3 · Execution Detail

Microsoft Word

Document-level capture

Task Mining captures the specific file open during work — file names often encode case context.

file V-Narrative 549550 (English) (26643764_0_1_347593) 354×

file V-Narrative 523977 (English) (26234010_0_2_341403) 116×

file V-Narrative 549428 (English) (26642399_0_2_349187) 88×

file V-Narrative 531855 (English) (26362027_0_1_346062) 84×

file Narrative 481786 (English) 84×

file V-Narrative 524568 (English) (26241711_0_2_343144) 34×

file V-Narrative 540056 (English) (26485638_0_3_344672) 31×

file V-Narrative 533936 (English) (26397592_0_1_338360) 25×

Level 3 · Execution Detail

Microsoft Teams

App-level capture

Only application name and dwell time available. Direct instrumentation would unlock deeper detail.

293 granular events captured, but no DOM, document, or message-level metadata is available for this application. Instrumenting this app directly with Pyze would unlock element-level detail.

Standard Flow

Representative application journey — the dominant app-to-app transitions observed in the pilot data. Tier badge shows how deeply Pyze captures each application.

Veeva Safety

DOM

11,601 events · 550 cases

Microsoft Word

Document

665 events · 129 cases

Veeva Safety

DOM

11,601 events · 550 cases

Phobos

DOM

1,460 events · 156 cases

Veeva Safety

DOM

11,601 events · 550 cases

Dominant transitions chained from real observations. Specific case journeys appear as variants below. Use the Step Journey above to drill into any task.

Step Variants

Distinct task-sequence patterns observed across cases in this step. Click a variant on the left to see its task sequence, business impact, and recommended remediations on the right.

Reference Narrative (step-level) Reference Path 165 cases · 10 analysts

Task Sequence

1. Select Narrative Template

Automated Explanation rule-based

This is the mined reference path — 165 instances across 10 analysts follow this exact sequence. Treat as the proposed Standard Operating Procedure for Narrative (step-level).

Reviewer Sweep Reviewer Verification Pass 127 cases · 11 analysts

Task Sequence

1. Open Narrative Workspace

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 7.3 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 11 analysts across 127 cases.

Time

946 hrs/yr

Cost

$70,950/yr

At 1,000

55,647 hrs/yr

Reviewer Sweep Reviewer Verification Pass 18 cases · 9 analysts

Task Sequence

1. Select Narrative Template → 2. Open Narrative Workspace

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 8.1 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 9 analysts across 18 cases.

Time

396 hrs/yr

Cost

$29,700/yr

At 1,000

23,294 hrs/yr

Backfill Loop Mid-Flow Data Backfill 14 cases · 1 analysts · 4× same-task rework

Task Sequence

1. Select Narrative Template → 2. Open Narrative Document → 3. Select Narrative Template

Automated Explanation rule-based

Backfill Loop — instances revisit the same activity 4.0× on average, consistent with mid-flow detours to fetch or enter data that wasn't ready at the right step. Ordering issue, not an app issue — 1 analysts across 14 cases. Suggests upstream data readiness or UX sequencing is at fault.

Time

15 hrs/yr

Cost

$1,125/yr

At 1,000

882 hrs/yr

Reviewer Sweep Reviewer Verification Pass 13 cases · 8 analysts

Task Sequence

1. Open Narrative Workspace → 2. Select Narrative Template

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 7.2 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 8 analysts across 13 cases.

Time

4 hrs/yr

Cost

$300/yr

At 1,000

235 hrs/yr

Reviewer Sweep Reviewer Verification Pass 13 cases · 2 analysts · 2× same-task rework

Task Sequence

1. Open Narrative Workspace → 2. Open Narrative Document → 3. Open Narrative Workspace → 4. Open Narrative Document

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 9.3 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 2 analysts across 13 cases.

Time

6 hrs/yr

Cost

$450/yr

At 1,000

353 hrs/yr

Reviewer Sweep Reviewer Verification Pass 10 cases · 6 analysts

Task Sequence

1. Review Narrative

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 4.1 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 6 analysts across 10 cases.

Reviewer Sweep Reviewer Verification Pass 8 cases · 2 analysts · 2× same-task rework

Task Sequence

1. Select Narrative Template → 2. Open Narrative Document

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 9.4 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 2 analysts across 8 cases.

Time

2 hrs/yr

Cost

$150/yr

At 1,000

118 hrs/yr

Reviewer Sweep Reviewer Verification Pass 8 cases · 3 analysts

Task Sequence

1. Review Narrative → 2. Open Narrative Workspace

Automated Explanation rule-based

Reviewer Sweep — a viewing-heavy pattern with 18.2 avg sub-steps and zero edits captured. Consistent with a downstream reviewer re-opening the task to verify earlier work rather than progressing new data. 3 analysts across 8 cases.

Time

25 hrs/yr

Cost

$1,875/yr

At 1,000

1,471 hrs/yr

📄

Detailed Step Reference

Full DOM element detail + case evidence for every step (5 total) — mostly for BA and RPA-developer use

Open Narrative Workspace

Open Narrative Section None

764 observations 319 cases 14 analysts Median 24.0s 100% high confidence

Select Narrative Template

Select Narrative None

595 observations 318 cases 14 analysts Median 28.0s 100% high confidence

Open Narrative Document

None

318 observations 143 cases 8 analysts Median 96.0s 100% high confidence

Generate Narrative (AI-assisted)

Generate Narrative from Outline None

63 observations 55 cases 9 analysts Median 32.0s 100% high confidence

Review Narrative

View Narrative None

62 observations 56 cases 10 analysts Median 60.0s 100% high confidence

Top Application Transitions

Top app-to-app hops observed during this stage. Each pair is a single swivel-chair move, ranked by frequency.

Rank	From App		To App	Cases	Frequency	Share
1	Microsoft Word Document	→	Veeva Safety DOM	121	364	22.1%
2	Veeva Safety DOM	→	Microsoft Word Document	117	309	18.8%
3	Phobos DOM	→	Veeva Safety DOM	146	288	17.5%
4	Veeva Safety DOM	→	Phobos DOM	143	284	17.3%
5	Acrobat App	→	Veeva Safety DOM	74	202	12.3%
6	Veeva Safety DOM	→	Acrobat App	76	197	12.0%

Path Variants

Top 10 full application journeys observed through this stage (consecutive duplicates collapsed). Shows the actual end-to-end paths cases take — not just single transitions.

Rank	Hops	Application sequence	Cases	Share of cases
1	1	Veeva Safety DOM	308	56.0%
2	3	Veeva Safety DOM → Phobos DOM → Veeva Safety DOM	30	5.5%
3	5	Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM	12	2.2%
4	7	Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM	6	1.1%
5	3	Veeva Safety DOM → Quality Vault DOM → Veeva Safety DOM	4	0.7%
6	9	Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM	3	0.5%
7	3	Veeva Safety DOM → Microsoft Teams App → Veeva Safety DOM	3	0.5%
8	11	Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM → Phobos DOM → Veeva Safety DOM	3	0.5%
9	3	Veeva Safety DOM → Notepad App → Veeva Safety DOM	3	0.5%
10	3	Veeva Safety DOM → Acrobat App → Veeva Safety DOM	2	0.4%

A "1-hop" path means the case stayed in a single application for this stage. Longer paths indicate cross-app work — those are the ones to investigate for swivel-chair reduction opportunities.

Applications Footprint

Where work actually happens during this stage. Cross-app time flags potential integration or UX opportunities.

Application	Capture tier	Events	Cases touched	Avg dwell per event	Total minutes in stage
Veeva Safety	DOM	2,211	550	0.56s	20.6 min

Drift Monitoring

Compares observed behavior against the published SOP. Drift signals feed compliance events and trigger automatic SOP refresh.

Drift level

High

Top path concentration

15.4%

Distinct transitions

133

Emerging steps

Unstable steps

Rare apps

Uneven adoption apps

Fragmentation signal This stage shows 133 distinct transitions with the top path at only 15.4% — a classic fragmentation signal. Monitor for continued drift.

Emerging activities (not in standard flow)

Activity	Frequency	Cases	Action
Preview Japanese Narrative	57	48	Candidate for SOP inclusion · Promote to step →
Preview English Narrative	48	37	Candidate for SOP inclusion · Promote to step →
Update Case Narrative	38	31	Candidate for SOP inclusion · Promote to step →
Select Spontaneous Narrative	23	23	Candidate for SOP inclusion · Promote to step →
View Site Narrative	19	17	Candidate for SOP inclusion · Promote to step →

Rare apps (used on <15% of cases)

Application	Capture tier	Cases using	Events	% of cases	Action
Acrobat	App	81	424	14.7%	Shadow tool signal · Flag for investigation →
Microsoft Outlook	Message	70	138	12.7%	Shadow tool signal · Flag for investigation →
Microsoft Excel	Document	65	178	11.8%	Shadow tool signal · Flag for investigation →
collaboration.merck.com	App	44	72	8.0%	Shadow tool signal · Flag for investigation →
ONENOTE	Document	28	46	5.1%	Shadow tool signal · Flag for investigation →
Notepad	App	24	39	4.4%	Shadow tool signal · Flag for investigation →
gpteal.merck.com	DOM	20	37	3.6%	Shadow tool signal · Flag for investigation →
olk	Message	14	21	2.5%	Shadow tool signal · Flag for investigation →

Uneven adoption apps (used by minority of analysts)

Application	Analysts using	Total events	Adoption	Action
Acrobat	6	424	38.0%	Training or knowledge-transfer opportunity · Review users →
Microsoft Outlook	6	138	38.0%	Training or knowledge-transfer opportunity · Review users →
collaboration.merck.com	6	72	38.0%	Training or knowledge-transfer opportunity · Review users →
apps.powerapps.com	4	48	25.0%	Training or knowledge-transfer opportunity · Review users →
ONENOTE	4	46	25.0%	Training or knowledge-transfer opportunity · Review users →
Notepad	1	39	6.0%	Training or knowledge-transfer opportunity · Review users →

Continuous monitoring active This SOP reflects observed behavior through the 19-day pilot. In production, drift signals are checked hourly and published SOPs auto-refresh when consensus shifts. Drift events appear in the admin audit log with a diff against the previous version.

Narrative

Tasks in this Step

Publish to QualityDocs

Step Journey

Standard Flow

Step Variants

Open Narrative Workspace

Select Narrative Template

Open Narrative Document

Generate Narrative (AI-assisted)

Review Narrative

Top Application Transitions

Path Variants

Applications Footprint

Drift Monitoring

Emerging activities (not in standard flow)

Rare apps (used on <15% of cases)

Uneven adoption apps (used by minority of analysts)