A claim is a story told by documents that rarely sit together. Ask the whole story.
"What happened, when, according to whom, and what in the file proves it?" docenta reconstructs the answer across policy, first notice, adjuster notes, correspondence, estimates, scans and the time and place inside the photos, with exact receipts and explicit gaps. Your claim file stays on your machine.
The earliest indexed report of visible mold is an email at 9:12 on February 14, one day after the earliest damage photograph; six photos were attached. The policy in the claim folder carries a $10,000 fungi sublimit on page 12. The collection does not establish whether the mold resulted from the covered water event: the mitigation report that would support that causal link is not present.
An illustration of the shape of an answer. The claim is made up; the behavior, timeline reconstruction from image metadata, policy retrieval, the contradiction and the disciplined non-inference, is docenta's.
You remember the meaning, the date, the argument. Ask that.
Timeline
Build the claim timeline from first notice through the latest supplement; cite every event.
When was the first mention of mold, water intrusion or roof damage, and by whom?
What happened between inspection and the denial, approval or payment?
Which version
Which policy version was in force on the date of loss, and what exact sublimit or exclusion language applies?
Which estimate is the latest, and which earlier numbers does it replace?
Photos as evidence
Which photos were actually taken before mitigation began, and what do their timestamps and locations show?
Which photographs have no usable timestamp or location metadata?
Contradiction
What did the insured report first, and do later descriptions conflict with it?
Are there different loss dates in the FNOL, the mail, the photos, the notes and the invoices?
Evidence and gaps
What evidence supports each disputed line item, and what is still missing?
Which request for documents is still unanswered?
What facts in the claim summary are not supported by the indexed file?
The policy and the notice.
The policy and its endorsements, the first notice of loss, the forms; each version kept with its date so "in force on the date of loss" is a question the file can answer.
Notes, mail and estimates.
Adjuster notes, correspondence with attachments in order, the carrier estimate and every supplement, invoices tied to the message that discussed them.
Photos and scans.
Image metadata (time, location) read and kept on the timeline; scanned pages and PDFs with OCR, and the extraction warned about when the text order is unreliable.
The same four-part contract, every time.
Exact source.
The file, the page, the message, the span. Not "sources": the line you can open and read.
Order and version.
When it was said, which version it was, what replaced what. A three-event timeline beats another paragraph.
Conflicts.
Two sources that disagree are shown side by side, never smoothed into one confident sentence.
Gaps.
"Not established by this collection" is an answer, printed as one. It is how you know the rest is solid.
The evidence path.
What was searched, which lanes answered, what was dropped and why, how many tokens the agent was handed. Inspectable, every time.
Boundedness is the feature.
Most AI marketing hides uncertainty. docenta prints it. The lines below are not disclaimers; they are what a trustworthy answer looks like.
A timestamp is not causation. A photo proves when and where it was taken. Whether the condition it shows was caused by the covered event is a finding the file must contain, or the answer says it does not.
First reported is not first occurred. The earliest indexed mention is dated and cited as such; the answer never turns it into the date of loss.
Missing documents are named. The mitigation report, the endorsement schedule, the signed authorization: what a reviewer would need and the file lacks is listed, not glossed.
Nothing decides coverage. docenta assembles the record with receipts; the coverage position is yours.
Back to the collections · also for contractors, lawyers, insurance, home, mail, developers, researchers · public collections, live: Judaism, Christianity, Islam, faith, sleep · in preparation: diabetes · the handbook
Ask the sensitive question without first uploading the sensitive file.
The claim file, the policy and the insured's photos stay on your machine: ingest, index, embeddings and answers all run there. Originals are never copied; only searchable derivatives are stored. Your agent gets a small evidence pack, a few kilobytes with the receipts, not your archive. Zero uploads, zero telemetry, and the token budget is printed on every answer.
It gets it over MCP, the connector standard Claude, ChatGPT and most coding assistants speak, from a server that runs on your own machine: the agent talks to a local process, not to a URL in someone's cloud. The only remote MCP addresses on this site are the public doors, sealed collections hosted on purpose; your own collection never has one. Deterministic layers, no learned model in the retrieval loop, bit-stable re-runs.
What is one question your files should be able to answer today?
docenta is a commercial product of SKY, LLC, now in private preview. Tell us the question; we will tell you when the door opens, and the questions we hear shape what we build.
No account and no newsletter: you confirm your address by mail, then we write back when the door opens. Prefer mail? insurance@docenta.ai