Grade Every
Lead Before
Hang-Up

During the kickoff call we capture your firm's intake priorities directly from your intake lead — case type, jurisdiction, severity bands, conflict rules, and the economic threshold below which you wouldn't take a matter. Those criteria become the rubric Claire grades against. We then calibrate by replaying your last 30 days of intake calls through the rubric until your team agrees with the grades. Tuning continues from there: every time your coordinator overrides a grade, the model updates.

Briefs ship to SMS, Slack, Microsoft Teams, email, and directly into your practice management CRM (Clio, MyCase, Filevine, Lawmatics, PracticePanther, HubSpot, and 60+ others). You can route different grades to different channels — A leads SMS the on-call attorney, all four grades land in the CRM as a structured record.

Each grade triggers a different cadence drafted in your firm's voice. A leads get attorney-callback prompts and retainer dispatch where pre-cleared. B leads get a consult-booking SMS, 24-hour reminder, and 72-hour second attempt. C leads get a single nurture message and CRM-only tracking. D leads get a polite referral or decline. Your intake lead approves the templates during onboarding; copy can be updated anytime.

Yes. ClaireAI runs on SOC 2 Type II infrastructure with HIPAA-aligned controls (our own SOC 2 Type II audit is in progress). Lead IQ data is encrypted with 256-bit AES at rest and in transit, stored in an isolated tenant per firm, and never used to train cross-customer models. Call recordings and transcripts are retained per your firm's retention policy.

Yes. The A/B/C/D rubric is firm-specific by design — a criminal defense firm calibrates around charge severity, jurisdiction, in-custody status, and prior counsel; a family law firm calibrates around custody factors, UCCJEA flags, and protective-order urgency. The same Lead IQ engine runs all three practice areas because the criteria layer is yours, not ours.

Traditional CRM lead scoring runs after the lead is written, based on form fields and behavioral data. Lead IQ runs on the call audio in real time — Claire grades the prospect from the conversation itself, not from a form. The brief includes case intelligence from similar prior calls, which is something generic CRM scoring can't produce because it has no access to call content.

Most firms are live in under a week: kickoff call (60 minutes), criteria definition (24-hour turnaround on our end), calibration against your recent calls (2-3 days), then go-live. The integration is included on every ClaireAI plan with no per-grade or per-brief surcharge.

Every grade is a recommendation, not a verdict. Partners and intake managers can override any A–D grade in one click from the brief — your override becomes a training signal, and Lead IQ recalibrates against your firm's rubric within the next intake batch. Statute-of-limitations edge cases (incident within 90 days of the SoL, discovery-rule jurisdictions, government claim deadlines) are hardcoded to bypass auto-rejection and route to human review regardless of score, so a wrong "D" never costs you a seven-figure case.

Yes. Every brief stores the full decision trail — the rubric criteria that fired, the call transcript moments that triggered them, the score weight per criterion, and any human override. Export the audit log to PDF, CSV, or push it into your matter file in Filevine, Litify, or Lawmatics. When a chiropractor referral source pushes back on a low grade, you can show them exactly what the intake said in 30 seconds.

Industry average lead-to-client conversion is 25–35%; elite firms push 55%+. Lead IQ lifts sign rates two ways: by routing A-grade calls to a live attorney within 5 minutes (which Forrester/InsideSales data shows lifts conversion roughly 300%), and by keeping B and C leads warm through grade-specific drips instead of letting them die in voicemail purgatory. Firms running Lead IQ for 90 days typically see sign-rate improvements of 8–15 percentage points on the same lead volume.

Yes. Lead IQ collects express written consent during the intake call ("Is it OK if we follow up by text?"), logs the timestamp and recording for proof, honors STOP and UNSUBSCRIBE within seconds, and enforces 8 a.m.–9 p.m. local quiet hours per CTIA rules. All campaign numbers are pre-registered under 10DLC with carrier-approved use cases. Opt-out, consent-revocation, and the FCC's new 10-day revocation rule are all built in — your firm never has to wonder if a $500–$1,500-per-message TCPA exposure is sitting in your drip queue.

Yes. Each practice group can define its own brief template — PI partners typically want injury severity, treatment status, liability facts, insurance limits, and prior-attorney flag; family-law partners want asset indicators, conflict level, custody complexity, and protective-order overlap; criminal-defense wants charge severity, custody status, prior record, and bond posture. Rubric criteria and brief fields are configured in the same place, so your "A lead" definition and your brief structure stay in sync.

Briefs generated during pre-engagement intake live in a gray zone — they're typically not protected by attorney-client privilege until representation begins, and AI-generated summaries have been held discoverable in cases like US v. Heppner (2026). Lead IQ defaults to storing briefs inside your matter management system (Filevine, Litify, Lawmatics) under attorney direction, which is the configuration NYC Bar Formal Opinion 2025-6 endorses for work-product protection. We never train shared models on your call data, and briefs are encrypted at rest in your dedicated tenant.

Yes. Lead IQ syncs bi-directionally with Filevine, Litify, Lawmatics (QualifyAI fields), CallRail tags, and any system with a public API or Zapier connection. Your matter-type taxonomy, case-value fields, and historical settlement data feed the rubric, and signed and declined outcomes flow back as training signals. When you adjust your case acceptance criteria in Filevine, the rubric inherits the change at next sync — no parallel spreadsheet to maintain.

A 12-minute call recording takes 12 minutes (or 4 minutes at 3x) to review; Claire's Brief takes 30 seconds. The brief is structured — injury, liability, damages, red flags, recommended action — not a wall of text. It also extracts what a human reviewer typically misses on a fast listen: missing fact-pattern elements, inconsistent statements, statute-of-limitations math, and prior-counsel disclosures. Recording is for compliance and disputes; the brief is for triage. You get both.

Lead IQ tracks every "D"-graded lead for 180 days post-intake and flags any that the firm later signs as a "false reject." Our published target is under 2% false-reject rate against firms' own historical decisions, measured monthly and visible in your admin dashboard. If false-rejects spike, the rubric is auto-flagged for review and we'll schedule a calibration session before the next billing cycle.

Every campaign is voice-tuned to your firm — we ingest your existing email and SMS templates, your website copy, attorney bios, and three sample sent messages, then generate drips that match your tone (folksy plaintiff vs. white-shoe defense vs. boutique family). Partners review and approve the first batch before anything sends. Open rates on tuned legal SMS campaigns typically run 95%+, and email open rates beat the 22–25% legal-industry average because grade-segmented messages are relevant to where the lead is.

Multi-defendant criminal cases (co-defendant conflicts, severance issues), parental kidnapping or domestic-violence overlap in family law, mass-tort plaintiffs, and class-eligible cases all trigger a "complex" flag that overrides single-grade output and routes the brief straight to a partner with the complexity reasoning shown. We don't try to grade "is this a multi-defendant murder-1 with a snitch problem" on an A–D scale — that's a partner conversation, and Lead IQ knows it.

Lead IQ auto-retrains every 14 days against your signed and declined outcomes — the same cadence Microsoft Dynamics 365 uses for its predictive lead score model, the established industry standard. You can also run rubric changes as A/B experiments: split incoming calls 50/50 between current and proposed rubrics, watch sign-rate and revenue-per-lead diverge for two weeks, then commit the winner. No "set it and forget it" — and no flying blind on changes.