underwriting
Multi-Turn-Insurance-Underwriting
Dataset Card for Multi-Turn-Insurance-Underwriting
Dataset Summary
This dataset includes sample traces and associated metadata from multi-turn interactions between a commercial underwriter and AI assistant. We built the system in langgraph with model context protocol and ReAct agents. In each sample, the underwriter has a specific task to solve related to a recent application for insurance by a small business. We created a diverse sample dataset covering 6 distinct types… See the full description on the dataset page: https://huggingface.co/datasets/snorkelai/Multi-Turn-Insurance-Underwriting.Multi-Turn-Insurance-Underwriting-Code-Gen
Dataset Card for Multi-Turn-Insurance-Underwriting-Code-Gen
This dataset is a variant of the Multi-Turn-Insurance-Underwriting dataset, in which models do not get access to any tools except a code interpreter and a pointer to the relevant file system.
This helps us analyze how well models explore their environments.
Environment Creation
This diagram shows the architecture of how we create the dataset, with assistant responses interleaved with questions, ending with a… See the full description on the dataset page: https://huggingface.co/datasets/snorkelai/Multi-Turn-Insurance-Underwriting-Code-Gen.credit-underwriting-preview
Credit Underwriting Benchmark
A 1387-task RL environment for financial document understanding. Agents are provided with business documents (bank statements), and asked
to assess the risk associated by using a context-rich classification taxonomy for each transaction . The agent's output is programmatically graded
against the ground truth labelled by domain epxerts with over 10+ years of experience in the risk assessment field.
Scale
1387 qualifying cases (Businesses… See the full description on the dataset page: https://huggingface.co/datasets/metaphilabs/credit-underwriting-preview.underwriting-dataset-blocks
Dataset Card for underwriting-dataset-blocks
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/JETech/underwriting-dataset-blocks/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/JETech/underwriting-dataset-blocks.insurance-underwriting-loss-coherence-risk-v0.1What this repo is for
Detect misalignment between underwriting assumptions and real losses.
Focus
pricing vs exposure
risk score vs claim trend
early signals before loss spikes
Why it matters
Insurers often discover pricing mistakes too late.
This dataset tests whether systems can detect coherence loss early.
gcc-insurance-underwriting-risk