Loss Run Analyzer
Ingests multi-year loss runs, normalizes carrier formats, and produces a frequency-severity narrative plus remediation talking points you can hand straight to an underwriter.
[ capabilities ]
- —Multi-carrier loss run normalization
- —Frequency and severity trend analysis
- —Underwriter-ready narrative export
[ core systems & integration architecture ]
Ingests loss runs from wherever they arrive, normalizes carrier formats, and attaches the analysis to the submission in your core system.
Guidewire PolicyCenter
Native REST APIAnalysis attached to the submission as a document
Applied Epic
Native REST APILoss summary written to the policy attachment store
Carrier loss-run portals
Batch & File DropScheduled retrieval and PDF normalization
Document storage
Event Stream / WebhookNew file in the watched folder triggers parsing
Sync model
Async — a five-year loss run analyzed in minutes
Standards spoken
Time to connect
Days
Enterprise guardrails
- —No customer data used to train models — zero-retention inference only
- —OAuth 2.0 / mTLS against your existing identity provider
- —Field-level audit trail on every value written back
- —Every extracted figure links back to the source page
[ mcp microservice interface ]
This module ships as an MCP microservice — lossrun-mcp. It consumes governed tools from the core-system adapters and publishes its own, so your assistant, orchestrator, or another module can call it directly. No credentials ever reach the model runtime, and every write lands in the human gate.
Consumes from core systems
- gw_extract_loss_run
- ivans_latest_download
- acord_parse_form
Provides to other agents
analyse_loss_historyreadFrequency, severity and trend from raw loss runs
compare_to_classreadBenchmark an account against its class code peers
[ implementation playbook ]
How an agency actually stands Loss Run Analyzer up — read-first, human-gated, and measured from day one.
Setup, in order
- 1
Gather a sample of loss runs from your top five carriers — every carrier formats them differently, and that chaos is exactly what the analyzer absorbs.
- 2
Define your standard output: the summary format your producers and underwriters actually want to read.
- 3
Connect the intake channel where loss runs arrive (usually email or a carrier portal download folder).
- 4
Set the analysis rules: lookback period, loss-ratio calculation, open-claim treatment, and large-loss thresholds.
- 5
Run parallel for a month: the analyzer summarizes every loss run, and a person verifies against their own reading.
- 6
Turn on the renewal trigger: loss runs for accounts 120 days out get analyzed automatically.
Agency use cases
Loss runs in minutes, not days
A carrier's 40-page loss-run PDF becomes a one-page summary — total incurred, loss ratio, open reserves, and the three losses that matter — before the producer finishes their coffee.
Renewal negotiation ammunition
The analyzer surfaces that the account's loss ratio improved for three straight years, giving the producer a documented case against a rate increase.
Prospect evaluation
On a new-business opportunity, the analyzer reads five years of loss runs and flags the frequency pattern that suggests a safety problem — before the agency commits capacity.
The first win to aim for
Loss-run turnaround from request to summary drops from days to same-day for every account in the pilot.
Guidance by agency role
Producers
Use the verified summary to frame renewal and prospect conversations, but inspect large losses and open reserves before presenting conclusions to a market.
Account managers
Confirm policy periods, claim status, and totals against the source document, then attach the approved summary to the renewal record.
Agency administrators
Set calculation conventions and verification samples, maintain carrier-format coverage, and monitor exception rates before automating renewal triggers.
[ built for ]
[ where this sits in the vendor market ]
We recommend buying first wherever the market already solves a problem well. Here is what you can buy instead of building this — and where this module earns its place beside what you buy.
Products that already cover this ground
Loss run and certificate automation with a data layer
Buy it when: Agencies that want loss runs pulled and structured rather than chased by email.
Where it falls short: Teams that need the pipeline inside their own warehouse and cannot accept a vendor-hosted copy.