AI Atlas - Daily updates on real-world AI deployments worldwide

ING Belgium's Site Reliability Engineering team faced severe class imbalance when training a machine learning model to predict whether planned IT changes would lead to Major, P1, or P2 incidents. By augmenting training data with synthetic replicas generated through DataCebo's Synthetic Data Vault (SDV) Enterprise, the team improved a LightGBM classifier's F1 score from 76% to 97% — a 21 percentage-point gain. The deployment covers hundreds of daily IT changes logged in ServiceNow and Jira Service Management.

Telepass (Italy — 75% of motorway tolls, 10M+ customers, 18 European countries) deployed Salesforce Agentforce in 6 weeks to autonomously handle 40,000 weekly customer conversations. The FAQ agent now resolves 87% of incoming inquiries on its own and is backed by Data 360 for knowledge grounding plus MuleSoft for SAP integration. A second internal Agentforce deployment has already cut service-rep average call handle time by 50%.

Sanofi's Quality Regulatory Science team is scaling Plai, an in-house AI quality-intelligence tool, from a deviation-trending pilot to a global deployment across multiple manufacturing sites. Built on the RAISE (Responsible AI at Sanofi Enterprise) governance framework, Plai accelerates investigations, generates root-cause hypotheses, and integrates QMI and QRE risk metrics to surface predictive insights before deviations occur.

University Hospital Tübingen (Uniklinik Tübingen) partnered with AI consultancy statworx on a pro-bono computer vision project that automatically determines the three-dimensional extent of brain tumors in MRI images. The model replaces the slow, manual drawing of tumor boundaries by radiologists, and handles heterogeneity across anatomical variants, tumor sizes, and imaging modalities while maintaining good segmentation accuracy. Per PD Dr. med. Florian Grimm, the team plans to extend the work via a third-party-funded collaboration project.

ThinkDeep AI's DeepBrain multi-agent platform has been deployed across the French Ministry of Economy and Finance and the French Ministry of Defense, saving €2M at scale across 10,000 Ministry of Finance employees, cutting document retrieval from two days to two minutes.

NHS England is rolling out Ambient Voice Technology (AVT) nationally under its 2026 Self-Certified Supplier Registry. The GOSH DRIVE London-wide trial of 17,000+ encounters recorded 51.7% less documentation time and 23.5% more patient interaction.

Alberta's Ministry of Technology and Innovation used Anthropic Claude Code with Opus and Sonnet models to scan 466 million lines of code across 1,280 provincial applications in 20 hours, surfacing security vulnerabilities that would have taken an estimated 6.5 years to find manually. The same internal team is now using Claude to fix vulnerabilities, modernize legacy systems (one ministry consolidating 185 legacy apps into 16), and run continuous security review agents. First-party source: Anthropic, published July 6, 2026 alongside Alberta's 21-document Velocity White Papers.

Allianz's Project Nemo, deployed in Australia in July 2025, is the insurer's first integrated agentic AI claims system, automating sub-AUD$500 food spoilage claims with seven specialised AI agents and reporting an 80% reduction in claim processing and settlement time.

Multiple Chinese manufacturers including Huaxiang Group's flange factory and Taisteel Stainless Steel (Taiyuan Iron & Steel's stainless steel unit) have deployed AI vision systems for inline quality inspection. Huaxiang's flange line achieves 100% online inspection with millisecond-level defect detection, while Taisteel's high-furnace inspection robots patrol steel-making facilities continuously to catch developing faults, reflecting China's broader industrial vision market shift toward data-driven, algorithm-decision production projected to exceed RMB 50 billion by 2028.

Foxconn deployed 2000 edge AI computing boxes across its Shenzhen Lighthouse factory, integrating AI vision models with production line equipment to detect product defects in real time. The system reduced defect rate from 0.8% to 0.15%, a 5x improvement, while enabling continuous quality inspection

Karnataka Power Transmission Corporation Limited (KPTCL) has deployed drones equipped with thermal imaging cameras and LiDAR sensors to inspect its high-voltage power transmission network across Karnataka. During a multi-week pilot, the drone fleet identified and rectified over 1.5 lakh (150,000) technical defects and reduced transmission-related outages by 85%, replacing manual tower climbing previously required in forests and other inaccessible terrain.

Autonomize AI — a healthcare-native AI orchestration vendor selected as a 2026 World Economic Forum Technology Pioneer — announced on July 15, 2026 the launch of Autonomize Genie AI, an AI agent that lets clinicians, operators, and IT teams design and refine production-ready healthcare workflows using natural language. The product targets utilization management, prior authorization, appeals, care coordination, and payer-provider workflows, generating auditable workflows from approved enterprise components rather than ungoverned code.

Ireland's distribution operator ESB Networks — serving 2.4 million customers — has launched a five-year program to digitally inspect up to 10,000 electricity grid structures with AI. Partnering with eSmart Systems, the utility combines UAV imagery, computer-vision defect detection, and human-in-the-loop validation to build a longitudinal asset record supporting risk-based replacement, predictive maintenance, and ESB Group's 2040 net-zero commitment.

AEP Ohio — a regulated electric utility serving approximately 1.5 million customers across 10,000 square miles and 47,000 distribution line miles in Ohio — piloted autonomous drone inspections across roughly 4% of its distribution system in 2025, collecting 400,000–500,000 aerial images. Partnering with Skydio for autonomous flight and Levitas for AI visual inspection, the program identified 150+ Tier-1 issues (predominantly thermal anomalies) and revealed the data bottleneck pushing AEP toward an integrated AI inspection workflow.

SK Telecom, Korea's #1 customer service provider for 27 consecutive years (NCSI), deployed Claude in Amazon Bedrock to power customer support for millions of Koreans. The deployment lifted in-call LLM response quality ratings by 34%, cut low-quality responses in SKT's Telco fine-tuned model by 68% versus the base, and reached ~89% of human agent quality in post-call processing — combining SKT's in-house RAG with Claude for end-to-end retrieval-augmented generation.

Rakuten, a Japanese technology company with 70+ businesses across e-commerce, travel, fintech, digital content, and communications, deployed Claude Code to compress time-to-market from 24 working days to 5 days (79% reduction) while sustaining 7 hours of autonomous coding on a complex vLLM refactor (12.5M lines, multiple languages) with 99.9% numerical accuracy. Claude Managed Agents were rolled out across product, sales, marketing, and finance within one week.

Bayer China built Dr. Bei, an LLM-based AI training system for pharmaceutical medical representatives running on Claude (Amazon Bedrock), going from concept to launch in 72 days and scaling to 1,000+ users within 6 months — saving 6,000–7,000 hours of training time per year and lifting 1-on-1 role-play coaching efficiency by over 50%. The system is now rolling out globally, with deployment underway in Europe and additional Asia regions.

Insilico Medicine's Rentosertib — the first antifibrotic small-molecule TNIK inhibitor, with both target and structure generated by generative AI — showed +98.4 mL mean change in forced vital capacity versus -20.3 mL on placebo at 12 weeks in the GENESIS-IPF Phase IIa trial (n=71), published in Nature Medicine. Patients on 60 mg QD not on concomitant antifibrotic therapy showed +187.8 mL. The first randomized clinical proof-of-concept that an AI-discovered target and AI-designed compound can deliver clinical efficacy in IPF.

Banco Santander announced June 22, 2026 that it has deployed 280 production AI agents, hit $40M ROI in Q1 2026, and extended AI access to all 185,000 employees worldwide; targeting $1.15B in business value by 2028.

NELFT (North East London NHS Foundation Trust) deployed EBO's AI Virtual Assistant named Harper across two early adopter services to digitise appointment management. The Trust, serving 4.3 million people across London boroughs (Barking & Dagenham, Havering, Redbridge, Waltham Forest) and Essex, Kent and Medway, achieved a 25% reduction in missed appointments and £169,135 in annual DNA savings. Patient engagement reached 43% out-of-hours and 91% patient satisfaction, validating conversational self-service as a viable NHS appointment booking model.