Eight checkpoints before a single dial.
Your firm's name never goes on a bad call.
TCPA liability is the first objection any attorney raises — and it should be. The penalties are statutory, the plaintiffs are organized, and the plaintiff bar has industrialized the process of identifying calling programs that cut corners.
This platform was designed for law firms. That means compliance is not a feature — it is the architecture. Before any production calls are placed in your state, we obtain written legal opinions from outside counsel confirming the compliance posture specific to your jurisdiction. That opinion is a hard condition of the service agreement, not a courtesy.
Eight Independent Checkpoints
Each checkpoint is independent. Failing any one of them blocks the lead — the remaining seven do not compensate or offset. This is not a scoring system. It is a gate system. A lead either passes all eight or it does not dial.
The FCC's 2023 one-to-one consent rule requires that consent explicitly name the calling party — not a lead aggregator, not a "marketing partners" catch-all. Every lead entering the platform is checked for a consent record that names your firm specifically. Any lead that cannot produce a verified, firm-specific consent record is blocked at the gate and logged. It never reaches the dial queue.
A TrustedForm certificate captures the consent session — exact form language, timestamp, IP address, and the claimant's interaction — at the moment of opt-in. The platform audits the captured form language against current FCC-23-107 disclosure requirements before accepting the lead. Certificates are attached to every lead record and retained for the full statutory period. Form language that fails the audit blocks the lead regardless of certificate validity.
Consent has a shelf life. A claimant who opted in six days ago is a different legal situation than one who opted in six hours ago — and plaintiff attorneys know the difference. Any lead whose consent timestamp is older than 72 hours is quarantined automatically. The system attempts re-consent via SMS before that lead is permitted to re-enter the dial queue. This is a hard block enforced in code, not a configurable soft warning that an operator can override.
The LITIGATOR_LIST is a database of phone numbers held by known professional TCPA plaintiffs — individuals who file serial TCPA suits as a revenue strategy. A single call to a number on this list can result in five-figure statutory damages. Before any number reaches the dialer, it is checked against the current LITIGATOR_LIST via DNC.com. Any match results in permanent suppression. This is the most consequential scrub in any outbound calling program, and it runs on every number, every time.
The National Do Not Call Registry is table stakes. What matters is that it is current, not cached. The platform scrubs against a live DNC feed — not a monthly snapshot — before each dial. Beyond the federal registry, applicable state DNC registries are checked for all states with their own maintained lists. Any number that returns a match is suppressed permanently from your firm's dial pool and flagged in the lead record. State registry coverage is updated as new state rules take effect.
Before the outbound AI call fires, the claimant receives an SMS from a registered 10DLC number — a short notification identifying your firm and providing a clear opt-out mechanism. This creates an additional documented consent touchpoint and gives the claimant an opportunity to opt out before any voice contact occurs. The 10DLC registration ensures deliverability and carrier compliance for the SMS leg. Opt-outs received at this stage are processed immediately and propagate across all downstream systems before the call is ever attempted.
At the opening of every AI-conducted call, the system identifies itself, names your firm, advises the claimant that the call is being recorded, and explicitly confirms the claimant's intent to engage with your firm about their Social Security disability claim. This exchange is captured in full audio and in the call transcript. The intent confirmation serves as an additional evidentiary layer — a documented, time-stamped, in-call record of the claimant's willing participation — beyond the initial consent certificate. Transcripts and recordings are retained per your state's applicable retention requirements.
Any opt-out signal — spoken verbally, triggered by keypress, or sent by reply SMS — is processed in real time. There is no batch processing window, no overnight suppression cycle, no delay of any kind. The number is suppressed across all active queues the moment the opt-out registers. The suppression event is logged with timestamp and mechanism. Any attempt to re-queue a suppressed number within the same session is blocked at the application layer. Opt-out records are retained indefinitely as part of the compliance record for your firm.
How This Is Permissible
The AI conducts intake — it gathers facts, asks structured questions, scores eligibility against defined criteria, and delivers information. It does not advise the claimant on legal strategy, predict hearing outcomes, evaluate the merits of a particular theory of disability, or make representations about the strength of the claimant's case.
The qualification interview is a structured data collection exercise. Every question asked is a question an intake coordinator would ask. The scoring criteria are thresholds your firm sets — the AI applies them, not interprets them. The attorney or representative reviews the output and makes the retention decision.
For attorney firms, the two-stage workflow is specifically designed around this boundary: the AI produces a written recommendation memo with supporting data, and the attorney makes the accept/decline decision. No client relationship forms, no legal advice is given, and no representation is implied until your firm affirmatively accepts the case through your normal processes.
State-specific written legal opinions from outside counsel are obtained before production calls begin in any jurisdiction. These opinions are specific to the authorized practice question — not generic technology opinions.
Every firm on the platform operates in a fully isolated model environment. Your call transcripts, claimant data, qualification outcomes, and attorney decisions are used only to refine your firm's model — they are never pooled into a shared training set, never used to improve scoring for a competitor firm, and never accessible to any other tenant.
The isolation is architectural, not policy-level. Tenant boundaries are enforced at the data layer. There is no configuration state that would allow cross-tenant data access.
Continuously refined by your data. Untouched by anyone else's.
SSDI qualification necessarily involves discussion of medical conditions, diagnoses, and treatment history. The platform treats all claimant data as Protected Health Information from first contact. Data in transit is encrypted via TLS 1.3. Data at rest is encrypted at the storage layer. Access controls are role-based and logged. Audit trails are retained for the full HIPAA-required period.
The HIPAA Authorization form is collected from the claimant during onboarding — walked through by the AI on the same call as retainer signing. Authorization records are stored in the claimant's file and transmitted to your case management system as part of the onboarding packet. Business Associate Agreements are executed with your firm before any PHI is processed.
Written outside counsel opinion before the first call. Every state. Hard contract condition.
Every compliance question answered before you sign.
Bring your outside counsel's questions. Bring your firm's skeptic. The 30-minute conversation is designed for exactly this. No proposal until you're satisfied with the compliance architecture.