Document Processing
CCC needed to train and iterate on large computer-vision models for insurance claims automation, but hybrid data movement, bursty GPU demand, and developer orchestration made it hard for data scientists to move models from experimentation to production quickly.
Reported impact
CCC reported a 2x speedup in running experiments, compressed some model-development cycles from months to days, and used the platform to launch CCC Estimate-STP, which can deliver detailed line-level collision estimates in seconds from photos instead of relying on workflows that can take several days.
CCC built a hybrid AI development pipeline around NVIDIA DGX Cloud, NVIDIA Base Command Platform, and DGX A100 systems to support the models behind CCC Estimate-STP. The company extracts structured and unstructured claims data from its production environment into sub-lakes, anonymizes it, and manages the resulting datasets in Base Command Platform so data scientists can schedule jobs consistently across on-premises and cloud GPU resources. Base Command's open API and Kubernetes compatibility let CCC integrate scheduling and resource management directly into its internal model-development workflow, while DGX Cloud provides 24x7 access to additional GPUs whenever training demand spikes. That platform supports the computer-vision and ensemble-model experiments used to turn mobile phone photos into line-level repair estimates in seconds, instead of leaving adjusters to rely on manual workflows that can take days.
Claims data is extracted from production into sub-lakes, anonymized and loaded into Base Command Platform, model-training jobs are scheduled across DGX A100 and DGX Cloud capacity, and the resulting computer-vision models feed CCC Estimate-STP so insurers can generate repair estimates from smartphone photos.
CCC Estimate (STP) uses AI, insurer-driven rules, and CCC's vast ecosystem to digitize the claims estimation process in seconds.
CCC Estimate-STP can deliver a detailed line-level estimation of a collision based on insurer rules in seconds from just a few pictures from a mobile phone.
Using Base Command Platform's dataset management and orchestration capabilities, our data scientists have reported a 2X speedup in running experiments.
Within DGX Cloud, we are using Base Command Platform's open API and compatibility with Kubernetes API to easily integrate it into our internal model development pipeline.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
NVIDIA Sample Sep 22, 2026 Sluicebox Sluicebox needed to turn fragmented bills of materials, supplier records, PDFs, CAD files, and emissions inventories into timely product carbon assessments, but manual life-cycle-analysis work was slow, expensive, and… 51-80% lower inference costs Lucy · NVIDIA Nemotron 3 Ultra · carbon graph NVIDIA Sample Sep 1, 2026 Onetera Onetera needed to help more than 90 municipalities turn fragmented regulations, zoning maps, permit rules, and legacy documents into fast resident-facing services without imposing heavy IT overhead or expensive manual… 1.7x increase in data throughput GPT OSS 20B · Government Intelligence Agent · LangGraph NVIDIA Sample Jun 26, 2026 MISA MISA needed to move from expensive, externally hosted LLM APIs to a self-hosted Vietnamese AI stack that could protect sensitive data, reduce latency and operating cost, and support many concurrent public-service and… $200,000 in monthly cost savings MISA AVA · NVIDIA A40 GPUs · NVIDIA DGX H200 