CaseCaliber
For attorneys · California plaintiff-side employment law

Built for California plaintiff employment firms.

You already know how to run an intake call. The problem isn't running them — it's running them yourself, between depositions and briefs and the third call of the morning. CaseCaliber sits on every intake call with you: the AI listens, surfaces what to ask next, and flags credibility issues live. When the intake's done, you trigger the analysis and get back a structured case file — statutes, damages range, SOL math, citations. Cases below your firm's threshold route through a Rule 1.5.1 network, launching with the founding cohort.

A morning, two ways

Same firm. Three intake calls before depositions at 10. Same partner.

Without CaseCaliber

  1. 8:00. First call — a wage-and-hour client. You're holding the intake form open in a tab and trying to type notes as she describes the schedule manipulation. You miss the moment she mentions a co-worker who saw the timecard edits.
  2. 9:05. Call done. You type up the notes (30 minutes), do the SOL math by hand (the client wasn't clear on the dates), and estimate damages from memory. It's 9:55.
  3. 9:58. You're due on the second call at 10. So are depositions. You pick: take the call, or be on time. Either way the third intake doesn't happen this morning.

With CaseCaliber

  1. 8:00. First call, with the AI on the line. It transcribes live, surfaces what to ask next, and flags the witness mention the moment it lands. You hang up at 8:38.
  2. 8:39. One click runs the analysis. By the time you've poured coffee, the case file is on your desk: §510 / §512 wage-and-hour, $180K–$320K damages range, SOL clean, two California Labor Commissioner precedent citations, witness flagged for follow-up.
  3. 9:50. Three calls done. Three case files on your desk. Depositions at 10 — on time, prepared, and the intake math, for once, is done.
Product 01 · AI intake & case analysis

What it actually looks like: the AI rides with you on every intake call — live transcription, dynamic question surfacing, branching when the matter shifts, credibility issues flagged as they happen. Multi-call support means one matter can span the calls it needs. When you're ready, trigger the analysis. The output is a structured case file built on a 258-field intake record.

[ UI mockup — annotated intake call screen ]

Mock depicts the call interface mid-call with three callouts: (1) Live transcript pane — client's words streaming in real time. (2) AI suggestion pane — next suggested question highlighted (e.g., 'Were timecards edited after submission?'), plus a 'Branch into wage-and-hour' prompt. (3) Credibility flag — inline alert ('Timeline mismatch — client said two weeks before on call 1, a month before here'). Below the call view, a clearly-positioned 'Complete intake → run analysis' button.

Intake engine phases
Phase What you see What you do
During the call Live transcript. Dynamic question surfacing across 32 baseline topics with conditional branches (discrimination, retaliation, whistleblower, wage/hour, high-value). Credibility flags on timeline inconsistencies, in real time. Conduct the call. Accept the suggested next question, ignore it, or go your own way — you're driving.
Between calls The case record so far: what's been covered, what's still open, which questions the system would prioritize on the next call. Schedule the follow-up call when the case needs one.
When the intake is done Case file with the AI's structured analysis: operative statutes with citations, damages range with itemized breakdown, SOL deadlines tracked in the case file, California precedent matches. Ratify, accept, decline, or route the case to the referral network.
Product 02 · Rule 1.5.1 referral network

Launching with the founding cohort. List a case across three visibility tiers — open, trusted, or curated/direct. Interested California plaintiff firms express interest during a window you set; you pick the best fit. The three Rule 1.5.1 documents — fee division, client disclosure, retainer — generate in mandatory sequence. Client PII transfers only after all three are signed.

Workflow built around the controlling California case law on Rule 1.5.1: Margolin v. Shemaria and Chambers v. Kay. Every step in the document sequence maps to a requirement from the rule itself.

[ UI mockup — annotated network listing screen ]

Mock depicts the network listing interface with five callouts: (1) Tier toggle (Open / Trusted / Curated/Direct) — showing which tier the case is currently published to. (2) Interest window timer — countdown showing time remaining for attorneys to express interest. (3) Interested-attorney panel — 3–4 attorneys who've expressed interest, with practice expertise, jurisdiction, recent matter examples. (4) Select CTA on each interested attorney. (5) PII-redacted case summary — shows case strengths, damages range, AI strength score; client name and employer name redacted.

  1. Fee Division Agreement

    Between you and the receiving firm. Sets the fee split, defines the joint-responsibility scope, signed by both attorneys. The next document unlocks only after this one is fully executed.

  2. Client Disclosure & Consent

    The client signs off on the proposed referral and fee arrangement. The consent language is explicit — not generic acknowledgment — the standard set by Margolin and Chambers. The client can decline at this stage; the case stays with the originating firm.

  3. Retainer Agreement

    Between the receiving firm and the client. Establishes the new representation. Can be generated from platform templates or uploaded by the receiving firm. On full execution, the case (including PII) transfers to the receiving firm; you retain a fee-tracking view.

Built on infrastructure your firm can trust

CaseCaliber is composed of named, audited infrastructure. No black box. Every integration is something your IT team — or your malpractice carrier — can verify.

  • Anthropic / Claude

    Primary AI analysis engine

  • Twilio

    Call orchestration and routing

  • Deepgram

    Real-time transcription

  • CourtListener

    California case law database

  • Trellis

    Legal data

  • DocuSeal

    Compliant e-signature for Rule 1.5.1 documents

Founding-cohort outcomes · what we're measuring

Targets, not invented retroactive stats. We will publish actuals from the founding-cohort firms 90 days after onboarding.

  • 15+ hours/month

    saved per partner on intake review

  • < 10 minutes

    from call-end to analyzed case file

  • Referral fees earned

    on declined cases — cohort actuals reported Q4 2026

Run the intake call with AI. Join the founding cohort.