VEGA- A Workbench for Insurance Claims Processing

VEGA- A Workbench for Insurance Claims Processing offers an autonomous AI PaaS ‘Libra’, with off the shelf modules for core insurance operations. The insurance claim workload automation module alleviates inefficiencies in the system by automating the entire process in near real time, from initial assessment and validation of a claim through the end decision and adjudication. Automated Claims Processing module uses ClaimsNet - fused NNs with deep layers of CNNs, RNNs and FNNs to learn the complex processes of ‘Claims’ directly from the historical data without any rules. The module, many times can match the process metrics of manual process with experts and sometime outperform the manual processing to automate ‘Pre-Authorization’, ‘Adjudication’ and scouting for ‘Suspicious transactions'. This AI-based application leverages the Intel® Optimized TensorFlow* framework running on the Intel® Xeon® Scalable processors, to secure performance boosts in inference workloads. Insurers can now promise faster claims settlements to users in near real time compared to current fragmented process with long turnaround times.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • Automated Underwriting and Pricing
    Automated Underwriting and Pricing evaluates insurance applications, determines risk, and generates personalized offers and premiums based on each applicant’s profile. It combines deep personalization with fast throughput.
  • Customer 360
    Customer 360 looks at a wide range of structured and unstructured customer data to provide a more complete and nuanced view of each individual. It can help to identify prospective customers and to create customized offers based on a better understand


WorldwideCross-IndustryFinance and InsuranceHealthcareAnomaly DetectionConversational Bots and Voice AgentsVideo Surveillance and AnalyticsCSP - Amazon Web ServicesCSP - Google CloudCSP - Microsoft AzureCSP - OtherOn-premise (Private Cloud, Other)Intel® Xeon® Scalable ProcessorIntel® Optimization for TensorFlow*Models cant be re-trained - Inference onlyLinux ProprietaryRNNDeep Learning