Live · refreshed daily from 58 vendors

Find AI deployments worth learning from.

For strategy, product, architecture and go-to-market teams: turn scattered customer stories from Microsoft, Google Cloud, OpenAI, IBM, AWS and more into comparable evidence about the problem solved, what was built, and whether the reported value is credible.

The decision problem

From marketing claims to usable customer evidence

Use the library to shortlist relevant deployments, inspect the architecture and outcome, then validate every important fact against its source before applying the pattern to your customer.

01

Who it is for

AI strategy, product, architecture, sales and research teams evaluating what has actually been deployed for customers like theirs.

02

How to use it

Filter by industry or use case, compare customer reference value, inspect the confidence breakdown, and open the cited sources before making a recommendation.

03

How to stay current

Subscribe to RSS for newly collected stories; the catalog refreshes daily and retains retrieval timestamps.

Latest

Newest customer use cases

Accenture Nov 12, 2026 Fincantieri Manufacturing Italy Fincantieri wanted to extend value beyond ship delivery by keeping vessels connected, improving operations over their long service lives, and using vessel data to support fuel efficiency, maintenance, and new digital… 100% real-time monitoring of core systems Predictive MaintenanceReal-time Analytics Fincantieri Ingenium · Navis Sapiens KPMG Nov 12, 2026 Undisclosed US retailer Retail United States A leading US retailer wanted to identify risks such as counterfeit products and litigation before costly incidents occurred. Existing manual reviews and fragmented public data could not reliably surface risks across… More than 55,000 data sources in KPMG Signals Repository Fraud DetectionRisk Management AutoML · BERT · BigQuery ServiceNow Modified · Nov 12, 2026 Medtronic Manufacturing Global Medtronic employees work under different country, contract, and regulatory arrangements, making a single HR knowledge base insufficient. The company needed a relevant, self-service way to answer routine HR questions and… 200 custom AI use cases in roadmap Customer Service AutomationIT Operations Automation HRmony · Moveworks · ServiceNow HR Service Delivery ServiceNow Modified · Nov 12, 2026 Georgia Secretary of State Public Sector Georgia, United States The Georgia Secretary of State needed reliable visibility into poll-pad patch levels and security across 159 counties, as well as a way to identify unknown assets and prioritize vulnerabilities with a small… 10,000 poll pads verified Security and Compliance Armis from ServiceNow · Mobile device management solution OpenAI Nov 12, 2026 Gilbert + Tobin Commercial and Professional Services Australia Gilbert + Tobin wanted to improve client service and employee productivity while enabling a law firm to use AI safely with sensitive work and confidential client information. 87% active usage among enabled ChatGPT users Document ProcessingEnterprise Knowledge Search ChatGPT Enterprise · ChatGPT Work · Codex OpenAI Nov 11, 2026 JioStar Media and Entertainment Asia-Pacific & Oceania Traditional recommendation systems were not enough to address the nuance of how viewers describe what they want to watch. JioStar needed fast, reliable, cost-conscious discovery for entertainment, live-sports moments,… 72.5 million concurrent live-sports viewers Codex · OpenAI API ServiceNow Modified · Nov 10, 2026 Copado Software / SaaS United States Copado's growing multi-cloud environment generated thousands of security alerts a day from siloed tools. The security team needed to correlate code, cloud, and infrastructure context, prioritize important… 17K findings remediated in a 30-day window Security and Compliance Armis from ServiceNow OpenAI Nov 10, 2026 Singapore Airlines Travel and Hospitality Asia-Pacific & Oceania Singapore Airlines wanted to improve customer support and travel discovery while giving employees governed AI tools for everyday work and software delivery. 2x increase in Kris customer satisfaction score Customer Service AutomationDocument Processing Flight Recommender · Jarvis · Kris Alibaba Cloud Nov 9, 2026 Yingkang Life (盈康一生) Healthcare China Yingkang Life needed to move agents from demonstrations into reliable production for a health ecosystem serving more than 13,000 employees. It needed stable quality, scale, cost control, auditability, and clear… 465 AI-native applications incubated Enterprise Knowledge SearchIT Operations Automation AgentCore · AgentLoop · Alibaba Cloud AI Gateway OpenAI Nov 9, 2026 Kuken Industries Manufacturing Asia-Pacific & Oceania Kuken's engineers needed to find the correct replacement-part drawing in a 70-year archive. Even after digitization, searches depended on knowing historic part names and numbers, consuming experienced engineers' time… 4.5 minutes to find a drawing, down from nearly an hour Document ProcessingEnterprise Knowledge Search Kuken drawing-search app · OpenAI API Salesforce Nov 9, 2026 City of Hope Healthcare United States City of Hope needed to let prospective patients begin intake after hours and on weekends without adding staff or taking patient-access representatives away from phone support. Existing chat in some markets depended on… 3.5% of total appointments were initiated through Agentforce in a 4-week period Customer Service Automation Agentforce · Health Cloud · Marketing Cloud KPMG Nov 6, 2026 Fortune 50 healthcare organization Healthcare Not documented A complex post-merger environment left the healthcare organization with an over-customized ServiceNow platform, technical debt, inconsistent processes, and multiple service systems that made support difficult to… 10,000 knowledge articles and 500 catalog items Customer Service AutomationIT Operations Automation ServiceNow · ServiceNow AI Control Tower · ServiceNow Virtual Agent

Auto-Generated Multi-Dimensional Intelligence

Where AI delivers quantified customer value

Quantified ROI benchmarks across 9,785 verified metrics

Median and 75th-percentile percentage improvements extracted automatically from customer-stated outcomes.

85% of stories report metrics

Cycle Time & Latency Reduction

Faster response, processing, turnaround, or deployment times.

952 stories
62% Median % Gain
88% P75 Top-Quartile
156 2x–100x Claims

Productivity & Time Saved

Engineering, analyst, or frontline staff hours saved and output gains.

463 stories
40% Median % Gain
70% P75 Top-Quartile
50 2x–100x Claims

Automation & Resolution Rate

Self-service containment, automated handling, and model precision/accuracy.

463 stories
80% Median % Gain
95% P75 Top-Quartile
11 2x–100x Claims

Cost & Infrastructure Efficiency

Direct operating cost, compute/storage TCO, and financial savings.

815 stories
50% Median % Gain
80% P75 Top-Quartile
32 2x–100x Claims

Revenue, Conversion & Experience Lift

Top-line growth, conversion rate, retention, CSAT, and NPS improvements.

314 stories
40% Median % Gain
81% P75 Top-Quartile
40 2x–100x Claims

Scale & Throughput Capacity

Volume of documents, transactions, queries, or users handled at scale.

802 stories
65% Median % Gain
87% P75 Top-Quartile
46 2x–100x Claims

6-Layer architecture components

10,633 mapped deployment components.

top product per layer
  • Ingest · 308 stories Top: MuleSoft (12), Pub/Sub (12), Spark Declarative Pipelines (11)
    364
  • Compute · 1,997 stories Top: Azure OpenAI Service (162), IBM watsonx.ai (101), Amazon Bedrock (95)
    3,473
  • Storage · 919 stories Top: BigQuery (143), MongoDB Atlas (90), Delta Lake (85)
    1,143
  • Serving · 1,773 stories Top: Power BI (63), Microsoft Teams (60), Databricks SQL (50)
    3,001
  • Orchestration · 1,235 stories Top: Gemini Enterprise Agent Platform (50), Gemini platform (48), Microsoft Dynamics 365 (37)
    1,624
  • Governance · 450 stories Top: Unity Catalog (109), IBM watsonx.governance (14), LangSmith (10)
    559

Top co-deployed product pairs

AI models and platforms most frequently combined in production.

product co-occurrence

Methodology

Transparent by design

Scores expose the strength and traceability of available evidence, not certainty or guaranteed ROI. A complete first-party source can score well without an artificial multi-source requirement.

  1. 01

    Track

    New customer stories are picked up from leading cloud and AI vendors every day, newest first.

  2. 02

    Trace

    Claims are linked to named sources, publication dates and retrieval timestamps so provenance stays visible.

  3. 03

    Score

    Evidence confidence and customer reference value are scored separately with a visible, deterministic rubric.

  4. 04

    Compare

    Use consistent problems, architectures, products and outcomes to find repeatable patterns—not just polished stories.

Find the strongest reference for your customer.

Search by vendor, industry, use case or product, then compare value, evidence confidence, architecture and source provenance side by side.