One workspace for the work around a case: deadlines, dockets, claims, settlement math, notice campaigns and public case pages. Every app hangs off the same matter, every finding carries its reasons, and every finding says what it could not check.
Give Bend the settlement date and the cascade fills in: business day math, holiday rollovers, court ordered flags. The calendar stops being a place mistakes hide.
Docket Monitor watches the courts you care about and mails the team the moment something lands, with the document named, not just noticed.
Allo Flow draws the distribution as a flow you can drag. Every change recomputes the waterfall and checks it against the rules while you build.
Casewatch publishes public case pages where every fact carries its source and a grade. What could not be verified says so, in public.
Links screens every arrival on a notice campaign, bots and duplicates out before anything is counted. The number that remains is the one you can declare.
Leak finds the first public surfacing of a guarded detail, proves the timestamp against the archive, and draws how the story spread outlet by outlet.
Bay gathers dockets, filings and public data sources into one library with one API. Ask it a case name and everything the record holds comes back together.
Docket Builder assembles a filing plan entry by entry, numbered and ordered the way the clerk will see it, so nothing files out of sequence.
Doc Review runs the privilege pass in the open: every document carries its state, and nothing leaves the room until the pass says it can.
Claim Flow tracks the pipeline from received to approved, and a deficiency letter goes out the day the gap is found rather than at the audit.
Notice plans derive from the class size on a diminishing returns curve. No tier claims one hundred percent, because that is not for sale, and the numbers survive a declaration.
Reminders watches the dates the other apps produce and mails the team while there is still time to act, not the morning it is due.
Risk Assessor scores each factor in the open instead of producing one oracular number, so the team can see exactly where the risk lives and push back.
Macro Dash draws prices, rates and risk straight from the federal series the analysts themselves cite, on one calm screen.
River Chat answers across everything the suite knows about a case, and every answer cites the app it came from, so nothing arrives on trust alone.
River Note holds the working record: every sentence can carry its citation, and the footnotes stay attached when the note travels.
Social Track keeps campaign delivery in one place, synced on a schedule, so the numbers in the room are the numbers on the platforms.
Social Ads turns a budget into a platform split and exports the order you actually place, so the plan on screen is the buy that runs.
White Label reskins the whole suite per organization: wordmark, colors and domain, with the same machinery underneath every brand.
Leak never treats a claim as a fact. An archive capture proves a page existed by that second, an outlet's own stamp is a claim a third party preserved, and a search date is only a claim. The Longview break reads proved public 09:11, not sometime that evening.
The Willowbrook sweep found nothing and recorded it as quiet, not clean: four outlets answered, seven could not be asked, one platform walled the query, and all eight gaps are named inside the finding. A clean answer that was really a blind spot is the worst lie software can tell, so it is the one this suite refuses.
Notice budgets follow a diminishing returns curve with a hard ceiling, so Maximum reach can price above Suggested and score beneath it. On a 60,000 person class, $3,500 outscores $5,500, and the model says so to your face.
Links classifies bots, infrastructure and duplicates out before a number exists, then prints what it excluded next to what it kept: 3,724 counted, 18 infrastructure and 3 duplicates set aside, every exclusion on its own line.
Casewatch pages carry a source and a grade on every fact. Verified means proved, reported means exactly that, and what could not be verified stands in public and admits it rather than quietly disappearing.
A claim starts at 1,000 points and loses them for named causes: the same email as another claim, no reachable contact, an amount far outside the case's range. Thirteen reason codes are registered today and one is already retired: it was written at 250 points, found to let a claim linked to another as one person still pass, and replaced rather than quietly reworded.
Determinations are append only and hash chained, and the chain is recomputed every time one is opened. A reviewer who disagrees issues a new finding that supersedes the old without erasing it, and the database refuses the edit outright. That is what makes how often review overturns the engine a query instead of a claim.
On a constructed population of 512 claims, identity resolution found all 52 planted duplicates across three filing patterns and linked none of the 15 households sharing an address or the 15 pairs of namesakes. No single weak signal can link a pair, so two people at one address stay two people until something else agrees.