Groundspeed

Loss Run Automation

Quickly see and understand loss histories

Insurance carriers receive thousands of loss runs each week, and underwriters invest hours into reviewing individual submissions for risk appetite and customer value.

Streamlining review and making underwriters more satisfied and more productive.

Groundspeed’s robust intelligence data processing capabilities

OCR & content preprocessing

Integrated state-of-the-art OCR and document preprocessing extracts the words, figures, and basic structural elements of documents into a consistent machine-readable representation.

Document classification

Artificial intelligence and machine learning models help classify and route submissions for efficient data extraction. They’re paired with systems and tools that rapidly identify low-confidence documents and improve the model.

Document parsing models

Document parsing models use each format’s known structure and schemes to automatically identify and classify the data points in any matching documents, as well as create an efficient pipeline for constructing new models.

Universal data schema

Our universal data schema encompasses the entire space of information relevant to commercial insurance systems. It allows us to accurately translate and structure extracted data points into a standard schema.

Standardization

Extensive business logic and predictive models infer data points that are implied but missing from the original documents. They also help us normalize the data field contents into standard data types and formats.

Quality assurance process

An automated data quality system with integrated human-in-the-loop quality control handles exceptions and ensures a high standard of reliable and complete data outputs.

We support any carrier layout for these documents

PDF Loss Reports
Excel/CSV Loss Reports

How it works

step 1

Document Ingestion

Insurance carriers submit documents to our system via API, SFTP, or email. Simple.

step 2

AI Data Labeling

Liberate underwriters from labeling. Our AI classifies, annotates, cleans, and structures data from those submission documents.

step 3

Automated QA

We run the captured data through a series of automated data quality and consistency checks.

step 4

Human Review

Our AI flags lower-confidence data and loops in a human to confirm accuracy. This process lets us take on more complex data sources than AI-only platforms and helps us continuously improve our technology.

step 5

Enrichments

We normalize the data and add insight. We fuse enrichments from our machine-learning models and from relevant third-party sources for the line of business into the data set.

step 6

Fast Delivery

Our technology delivers back accurate, normalized, and enriched data, typically in under two hours, via API, SFTP, or email. Making underwriting decisions quicker, easier, and more profitable.

Automation leads to revolutionary results

Interested in learning more about our Loss Run Automation product?