How the intelligence is built.
We go where the answers are: hard-to-find data from public and proprietary sources, refined by advanced machine learning into the patterns that drive value across the claims economy.
01 Hard-to-find data
DAIS goes after the data others overlook. We assemble court dockets and case filings, state insurance and regulatory filings, property and permit records, and our own proprietary datasets, reaching across the jurisdictions and sources that are difficult to compile by hand. Coverage spans more than a dozen states across the claims economy.
02 Machine learning & pattern detection
Advanced machine learning links, resolves, and analyzes the data at a scale manual review can't approach, surfacing the patterns that actually move a case: defendant and carrier posture, repeat-defendant behavior, settlement benchmarks, and exposure. The models do the heavy lifting so the signal rises to the top.
03 From data to a decisive brief
The output is a decision-ready intelligence brief, concise, sourced, and framed for the questions that matter in any claim. A clearer read on the other side and the market, so you negotiate and litigate from a more informed position.
04 Built responsibly
Intelligence is delivered in aggregate, anonymized form: a read on institutional conduct across the market, not records about individuals. DAIS describes patterns in carrier and builder behavior at the population level. DAIS Analytics, LLC is a data-analytics company; it does not provide legal advice or form an attorney-client relationship. How the intelligence informs a matter is always the attorney's judgment.
What data sources does DAIS Analytics use?
DAIS draws from public regulatory filings, state and federal records, court dockets, and proprietary data sources spanning more than a dozen states. The specific dataset catalog is available to Founding Members; detailed sourcing is also shared with investors.
Is the intelligence individualized or aggregate?
All DAIS intelligence is aggregate and anonymized at the market level. Outputs describe patterns across builder portfolios and carrier claim populations, not individual people. DAIS does not provide individualized records or personal data.
Does DAIS Analytics provide legal advice?
No. DAIS Analytics, LLC is a data-analytics company. It is not a law firm and does not provide legal advice or form an attorney-client relationship.
What machine learning does DAIS use?
Advanced AI and machine learning power multiple stages of the analytics pipeline, assembling fragmented records into coherent portfolios, surfacing behavioral patterns at scale, and producing the final intelligence brief. The technical methodology is shared with Founding Members and investors.
What jurisdictions does DAIS Analytics cover?
Carrier Intelligence covers more than a dozen states, shown in the map below. Builder Intelligence covers Florida, with additional contractor data linked from other states we cover.
- Carrier Behavior Index (CBI)
- A market-relative index that tracks how an insurance carrier's claims-handling conduct is trending relative to its peer cohort — across dimensions including claim-closure speed, litigation rate by claim type, and regulatory history. CBI is computed on a rolling basis from DAIS's underlying carrier propensity and conduct scores, and normalized by market size so carriers of different scale are directly comparable.
- Builder Risk Index (BRI)
- A composite score that measures a construction builder's defect exposure profile relative to its peer cohort, drawn from permit records, litigation history, and complaint data across the builder's full portfolio. The BRI reflects patterns across a builder's entire activity record — not just a single project — and is normalized by construction volume so builders of different scale are directly comparable.
- Entity Resolution
- The process of determining that two or more records from different data sources refer to the same real-world entity. In the DAIS context, entity resolution links a builder's subsidiaries, related companies, and predecessor entities into a single unified portfolio — making it possible to see a builder's full litigation and defect record across dozens of legal names, rather than treating each subsidiary as a separate and unrelated defendant.
- Aggregate Intelligence
- Intelligence derived from patterns across many cases, claims, or filings — rather than from individual records about specific people or individual claims. DAIS delivers aggregate intelligence exclusively. The output describes how carriers and builders behave at the market level — their patterns, postures, and trajectories across their books — not what happened in any particular claimant's case.
The analytics DAIS produces are not the output of a single model or a simple query. They are the result of a multi-stage pipeline that assembles, links, scores, and indexes millions of records before a single output is delivered to a subscriber.
Entity Resolution
DAIS uses probabilistic entity resolution to link thousands of legal entity names — subsidiaries, related parties, and predecessor entities — into unified enterprise profiles. The algorithm compares names, addresses, registered agents, and officer signatures across permit records, court filings, and licensing databases to determine whether two records refer to the same real-world enterprise. Precision and recall are validated against ground-truth sets built from known entity relationships. This is what makes it possible to see one builder's total litigation posture across dozens of subsidiary names, rather than treating each subsidiary as a separate defendant.
Propensity Scoring
Each carrier and builder is indexed against its peer cohort on multiple conduct dimensions: claim-closure speed, litigation rate by claim type, consent-order history, and financial-health trajectory. The index is normalized by market size and exposure so a small carrier with a high rate is distinguishable from a large carrier with a comparable raw count. Scores are updated on a rolling basis as new filings and disclosures are processed.
Training Data Scale
The models are trained and validated against millions of public filings across the active coverage states -- regulatory filings, court records, permit databases, and licensed data sources. Data is ingested on a continuous basis, with the scoring layer re-run quarterly or more frequently when a significant filing event occurs for a covered entity.
Named Scoring Indexes (BRI & CBI)
The Builder Risk Index and Carrier Behavior Index are DAIS's named scoring indexes. They are built on top of the multi-dimensional propensity and conduct scores from the prior stages — the Builder Risk Index as a composite exposure score, and the Carrier Behavior Index as a rolling, market-relative measure of how a carrier's conduct is trending against its peer cohort. Each output is a calibrated, market-relative index — not a raw count — that tells a practitioner or analyst where a carrier or builder stands relative to its peers on each conduct dimension.
If you’d like to walk through the methodology, Founding Members get a direct line to the team.