Centralized AI-Powered
Clinical Data Assessment
First R&D Accelerator Product deployed at scale — embraced by users and business sponsors
Why ClinShow Matters
A product-level overview for leadership — where we are, what we deliver, and where we go next.
🎯 The Vision
ClinShow is Sanofi's centralized, AI-powered clinical data assessment platform. It replaces fragmented legacy tools (SAS listings, static PDFs) with a single, real-time environment for clinical data review, validation, and predictive risk monitoring — deployed across 88 live studies and 800+ users.
💰 Business Impact
- €10M+/year efficiency savings from near-real-time data access and AI-powered assessment
- -2 weeks database lock acceleration on every study
- -25% queries to study sites through predictive KRIs, improving investigator relationships
🚀 Key Achievements (Q1 2026)
- Patient Profile released across all Pharma studies with queries integration and Snowflake migration
- Data Validation Agent approved by DRC and in limited release — first AI agent in clinical data ops
- Data Quality Prediction entering UAT — ML models predicting AE under/over-reporting
- 93.7% user activation rate and 83% satisfaction (NPS 39)
📊 User Engagement Highlights
- 93.7% activation — 759 of 810 granted users are active. Near-total adoption across all roles.
- 77.4K sessions this quarter — engagement nearly doubled between January and March (+85%).
- Patient Profile: instant traction. 52 events/user in its first month — 2nd most-used feature on day one.
- Review Tracker +492%. From 2 to 18 events/user in 3 months — review workflows are now core to daily usage.
- Data Managers are power users. 74 active users with the broadest feature footprint — only role using Patient Profile, Study Config, and Review Tracker.
🔬 Product Pipeline
ClinShow in Numbers
Key performance indicators showcasing adoption, speed, and user satisfaction across the platform.
Engagement Dashboard
Usage analytics from GA4 — Jan 27 to Apr 26, 2026.
Stacked by Therapeutic Area — Dec 2025 to Apr 2026
Event Volume
Total GA4 events across all studies
Studies with at least one event in the month
Study Reach
Number of active studies per month
Daily event count (subject_selected) from launch — Apr 13 to Apr 28, 2026
27 studies active on Patient Profile within 15 days of launch — immediate adoption across the portfolio.
Peak on day 2 (Apr 14) driven by EFC17504 power users. Usage then stabilizes at a healthy daily baseline.
Sustained engagement. No drop-off after initial spike — users come back daily, confirming the feature solves a real workflow need.
Daily Patient Profile Activity
subject_selected events — all studies combined
Data Managers dominate. 74 active users (38% of identified roles) with the highest feature breadth — only role using Patient Profile & Study Config.
Collaboration is passive. ~4,400 comment list views but only 257 comments created (5.8% conversion). Users read, rarely write.
Events per user vs active users — who are the power users?
Data Managers lead in reach. 74 active users — largest pool by far. High volume but moderate intensity per user.
SMMs are power users. Fewer users (20) but 4.7 sessions/user — deep, repeated engagement with clinical workflows.
GSOs punch above weight. Only 15 users but high event density — safety-focused roles drive disproportionate activity.
Activity vs Reach by Role
Bubble size = total events — top-right = most engaged
How deeply each role uses each feature
Data Listing is the workhorse. Highest events/user across almost all roles — the daily go-to view for clinical data.
Collaboration skews toward DMs. Data Managers generate most collaboration events. Other roles interact but create fewer comments.
Power BI adoption is uneven. Strong usage by DMs and GSMs, but low or zero for safety-focused roles — opportunity for targeted onboarding.
Average Events per Active User
Higher = deeper engagement with the feature
Which roles use which features — signals for targeted investment
Data Managers have broadest access. Only role touching Patient Profile, Study Config, and Review Tracker — widest feature footprint.
Core 3 features used by all. Comment List, Data Listing, and Power BI Reports are universal across every role.
Rave Queries is niche. Only DMs and SMMs actively use it — consider whether broader access would reduce query turnaround.
Role x Feature Heatmap
Active users shown — darker = more users
Our Personas
Who uses ClinShow and what activities they perform.
Ensures integrity, quality, and compliance of clinical trial data throughout the entire study lifecycle. Follows eCRF setup, supports database design, centralized monitoring activities, data cleaning, and study conduct until final database lock.
Provides medical oversight and expertise for clinical trials. Ensures scientific integrity, patient safety, and protocol compliance throughout the clinical trial process. Part of the Core Study Team.
Plays a crucial role in the planning, execution, and oversight of clinical trials. Assists the SMM on a day-to-day basis, provides support on operational activities, and helps with regulatory issues and ethics committees.
Specialized clinical operations professional responsible for centralized, risk-based oversight of clinical trial patient data without conducting physical site visits. Continuously analyzes data flowing from investigative sites to detect potential risks before they become critical. Works in RBM team pair with a Data Scientist (CM handles communication & interpretation; DS handles programming & technical). Future main user of the Clinshow predictive KRI module.
Monitors safety signals across studies. Reviews AE reports, assesses causality, and ensures timely reporting to regulatory authorities.
Orchestrates study operations globally. Tracks milestones, coordinates teams, monitors study progress via KRI dashboards.
Pharmacovigilance specialist supporting signal detection and benefit-risk assessments across the portfolio.
Manages collection, entry, and verification of clinical trial data. Works with EDC systems, supports audit readiness, and maintains data quality standards.
What's New
Recent milestones and releases across ClinShow.
Patient Profile LIVE
The reviewer opens the Safety Narrative, extracts every clinical fact (dates, diagnoses, symptoms, medications, lab results, procedures), then locates the matching CRF page and verifies the data is entered, consistent, and complete. When something is missing or inconsistent, a manual query is raised to the site.
Latest news
- Apr 13 — Released on all Pharma studies (10 stories shipped):
- Queries integration — Rave queries visible per record with detail drawer
- Navigation overhaul — Left menu rebuilt based on SAS PP structure
- Smart display — Empty tables hidden by default, search bar, adaptive column widths
- Domain enhancements — Smoking Habit, Medical History, Medications, Central Lab
- Feb 26 — Data migrated from Teradata to Snowflake. Patient Profile is now AI-ready.
- Next — Medical Review Agent: compare Safety Narrative vs Patient Profile data to auto-detect inconsistencies and suggest queries.
Data Validation Agent LIMITED RELEASE
The DV Agent automates this by running 28 CMS validation rules on AE & LB domains (representing 46% of total DV scope) across 3 PHARMA I&I pilot studies. The agent flags anomalies, suggests query text, and supports bulk-validation — all with a human-in-the-loop approval workflow.
Latest news
- Mar 25 — Limited release after full DRC approval (10 stories shipped):
- Anomaly views — List per rule, subject detail with AI analysis, affected subjects count
- Navigation & UX — Domain menu (renamed from "Anomalies"), table sorting, anchored guidelines panel (CMS, DRMP, Protocol)
- PP integration — Open anomaly directly on relevant Patient Profile domains (AE, LB)
- Rule enhancements — Enriched rules list, rule detail with AI card, Rule ID linked to CMS
- Now — SMEs labeling outputs (TP/FN/TN) to validate accuracy. User research ongoing to improve UX.
- Next — Roll out to all users on 3 pilot studies once scores validated. Target: -50% DV workload. Then scale to all data families.
Data Quality Prediction UAT
Predictive KRIs use ML models to detect issues before they materialize. The first use case, AE Under-Reporting (AEUR), detects sites/subjects with potentially under-reported adverse events. The model analyzes demography, labs, vital signs, medications, exposure, and medical history to predict expected AE patterns. A second use case, AE High-Reporting, detects over-reporting patterns (more AE dates, SOC variety, outcome variations than expected).
Latest news
- Apr 28 — Entering UAT phase. First use case: AE Reporting Predictions (under- and over-reporting).
- Next — Production on 2 pilot studies targeted end of May 2026. Target: -25% queries to study sites.
- Later — Expand to Missing Critical Data Prediction, Query Rate Prediction, and Data Entry Time Prediction. Goal: make prediction the standard for data monitoring across all studies.
Data Review Agent COMING NEXT
The Data Review Agent will go further by automatically detecting misalignments between collected data and the study protocol, surfacing findings to the right people, and enabling timely course-corrections.
Latest news
- Apr 26 — Solicited Events dashboard added for all Vaccine studies (study endpoints).
- Now — Backlog of 15 new dashboards + 15+ enhancements. Specific dashboards built by China DMP team.
- Next — AI agent MVP planned for Q3 2026. Built on Patient Profile, Snowflake, and Validation Agent patterns.
OKR Tree
From business goals to product metrics — how every initiative connects to value.
Objectives
Objectives
Q1
Indicators
Indicators
High Level Roadmap
Features grouped by value delivered — Q4 2025 to Q4 2026.
*Source: Jira CLINSHOW project, label "Roadmap" — features and timelines are working hypothesis and might evolve as we test and learn