Skip to content
Ishqkala

Compare · vs. Rule-based keyword and regex search

Ishqkala vs. Keyword Search Tools

Keyword search finds words. Subrogation lives in who did what to whom.

Start free trial Request free closed-claim audit

Keyword and regex tools scan claim notes for terms like "rear-ended", "contractor" or "recall" and raise a flag when one appears. They are cheap, fast and easy to audit. They also flag negated statements, miss paraphrases and misspellings, and cannot read scanned police reports. Ishqkala reads the full narrative and attachments, works out which party caused the loss, applies the state fault rule and scores each file by dollars and SOL urgency.

DimensionIshqkalaRule-based keyword and regex search
Matching methodReads sentences for who did what, to whom, and whenMatches listed words or patterns
NegationSeparates "OV ran the light" from "no other vehicle involved"Both lines can raise the same flag
Shorthand and typosHandles "OV", "clmt veh", "t-boned" and misspellingsOnly terms someone thought to list
Scanned documentsOCR and layout reading for police reports, origin reports and estimatesUsually limited to text fields in the claims system
Liability allocationEstimates the fault split and applies the state ruleNone; a hit is a hit
PrioritizationRecoverable dollars x win probability x days until SOLUnranked hit list, often sorted by date
False positivesLow-confidence files labeled as such, with reasoningGrow as the term list grows
SOL awarenessState SOL and municipal notice dates on every fileOutside the scope of a text match
Recall matchingModel and serial numbers checked against CPSC and NHTSA recallsSeparate manual lookup per file
MaintenanceUpdated centrally by our engineering and legal research teamsTerm lists tuned by hand in-house
TransparencySource sentences highlighted, rule applied statedVery clear: you see exactly which term matched
CostContingency, or platform fee banded by claim volumeOften already licensed or built in-house

Why keyword flags look good in a trial

The first week of a keyword trial finds real recoveries: the files where an adjuster typed "rear-ended by OV" in plain words. Then the term list grows to catch more, and the flags grow faster than the recoveries. Adjusters carrying 100 to 150 active claims learn to ignore them, and the tool goes quiet.

Where the recoveries actually sit

Over 80% of the opportunities Ishqkala finds are triggered by unstructured narrative text, and that text rarely uses the words on a list. "Insd states OV crossed center line", "FM traced origin to dishwasher control panel" and "plumber installed new supply line two weeks prior" each point to a liable third party. A term list catches them only if someone predicted the exact phrasing.

What changes for the subrogation desk

Instead of a long, unranked hit list, your team gets fewer referrals, each scored and backed by highlighted source sentences. Files the engine is unsure about are labeled that way, so reviewers spend their time where judgment is needed.

FAQ

Frequently asked questions

Yes. Customers often load their current term list so the first manifest can be compared side by side with existing flags. The comparison shows which flags were noise and which recoveries the list missed.

Each statement is read in context, so "no other vehicle involved" counts against a recovery and "insured believes OV may have been speeding" is weighed as an unconfirmed allegation. The reasoning shown with every score cites those sentences.

No. Ishqkala does zero AI model training on customer claim files or adjuster notes, and your data stays in US AWS regions.

Find the millions hiding in your closed claim files before the Statute of Limitations expires.

We'll scan a batch of 500 closed claims for free and deliver an itemized recovery manifest within 5 business days.

Request a free closed-claim audit Start 14-day free trial