Document Processing
AmTrust needed to turn broker-submitted Excess and Surplus policy requests into comprehensive quotes much faster, but its manual clearance process had to interpret unstructured email packages and attachments before underwriters could even begin tailoring a response.
Reported impact
AmTrust shortened its initial underwriting response from a multi-day manual process to same or next day turnaround, processed most submissions within hours, improved broker goodwill, and created the capacity to absorb higher quote volumes without expanding the old manual clearance workflow.
AmTrust worked with TCS to replace a manual Excess and Surplus clearance flow with a low-code, AI-based automation platform that ingests broker email submissions and their unstructured attachments, classifies each package, extracts policy-relevant fields, validates the captured data, and auto-routes exceptions for human review. TCS designed the solution with AmTrust's COO and IT team, mapping business rules and workflows so the normalized submission data could be formatted consistently and ingested into AmTrust's existing technology stack. Once the workflow completes, underwriters receive an account and shell quote ready for customization instead of starting from raw emails. The rollout covered all E&S products within 12 weeks and turned an expert-heavy intake process into a repeatable digital pipeline.
Broker partners send new-business submissions and attachments into the clearance process, where AI based classification extracts policy data from the unstructured files. Data validation checks the captured fields, the automation workflow routes straight-through cases or exceptions as needed, and the normalized package is ingested into AmTrust's technology stack so underwriters receive a ready account and shell quote.
"More than 80% of submissions are processed in less than four hours after the broker submits a request for a quote."
"The solution involves AI based classification, extraction of data from email attachments followed by data validation and automation of clearance process with provision to auto route cases that needed manual intervention."
"The solution was implemented within 12 weeks, followed by roll out to cover all E&S products."
"Using a low-code, AI-based automation platform, TCS transformed the entire clearance process"
"turn-around times for initial underwriting response have been reduced from a few days to same/next day"
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
OpenAI Sample Oct 16, 2026 Benefit Street Partners Benefit Street Partners needed to answer investment-committee follow-up questions faster, compare lender marks across shared loans, and review 300-page credit agreements for clauses that could shift value away from… 10–12 hours total. BDC mark tracker · Codex · Credit agreement analyzer OpenAI Sample Oct 16, 2026 Itaú Itaú needed to make financial guidance simpler and more personalized for millions of customers while also scaling document-heavy internal workflows, all under the governance, observability, and economic controls… 100% of the customer base is the goal by the end of 2026 Customer DNA · IARA · OpenAI models 
Google Cloud Sample Modified · Oct 9, 2026 Policybazaar Motor insurance options in India can be complex, which is why Policybazaar built its market-leading reputation by making the consumer journey entirely transparent and accessible. The digital platform empowers millions… Achieved 97% issuance within five minutes in PAYD using Gemini models Cloud Storage · Gemini 2.5 Flash · Gemini 2.5 Flash-Lite 
