Product

Built for the three people who run a collections floor.

The owner who operates alone, the supervisor who coaches the floor, and the agent who does this two hundred times a day. Every screen below is from the product design — labeled with what ships today and what is on the way.

For the owner

Self-serve from import to audit. No admin certification, no professional-services invoice.

Owner dashboard with connect rate, abandonment gauge, promises and live floor
Product preview of the dashboard shipping in v0.2.
Available today

Run the agency from one screen

Connect rate, abandonment against the 3% safe harbor (with the auto-throttle indicator), calls blocked by compliance with one-click audit links, promises captured in dollars — all pushed live. No BI-tool scavenger hunt, no CSV-by-email.

Campaign wizard final step: compliance preflight before activation
Product preview. Wizard, CSV mapper and engine gates ship today.
Available today

Create a campaign in three steps — not six pages of settings

Name and pacing, calling windows and caller IDs, then a compliance preflight that gates activation. Import accounts by CSV through a column mapper with dedupe and a per-row exceptions file. An owner does this alone, in minutes.

Penetration analytics funnel: attempted, reached, effective contact
Product preview. Penetration analytics and compliance-aware rewind ship in v0.2.
Available today

Penetration analytics, deduplicated per person

Unique accounts attempted, reached, and effectively contacted — the number your clients actually ask about — computed live from the call record, with drill-down to the individual attempts. Plus a filtered re-dial (rewind) that previews what compliance would block before you requeue anything.

Also for the owner: per-tenant API keys and a full REST + websocket API (the dashboard runs on the same API you get), seat-based billing with no usage surprises, and BYOC trunk setup — self-serve guide on Starter, assisted on Growth and Scale, white-glove included on Enterprise.

For the supervisor

Everything pushed, nothing polled — and evidence behind every number.

Live monitor: agent state grid, queue depth, abandonment gauge
Product preview. The supervisor audio controls are live in the API today; this monitor screen ships in v0.2.
Available today

A floor view with zero refresh buttons

Every agent's state and durations, calls in flight, callback queue depth and the abandonment gauge — pushed over websocket as they change. Listen, whisper, barge and take-over are live in the core API today.

Compliance audit trail with per-attempt frozen decisions
Product preview. The append-only evidence, the call log and the violations-prevented report ship today; this dedicated audit screen ships next.
Live in the core API

Answer any 'why did this dial?' in seconds

Every attempt carries its frozen compliance decision — frequency counter state, local-time window, suppression checks. Blocked attempts are first-class records, so your monthly compliance story is violations prevented, not violations logged.

For the agent

Keyboard-first wrap-up, a calling-window countdown in the header, and a workspace that never drops the call on navigation.

Agent wrap-up: typed promise-to-pay fields and callback scheduling
Typed PTP, promises and callbacks are live in the API and dashboard — and so is the in-browser agent workspace: one-click browser audio over WebRTC, or agents connect with any standard SIP softphone.
Available today

Wrap-up that produces clean data every time

Disposition required before the next call. Promise to Pay forces an amount and a date. Callbacks are scheduled with provenance and at least 10 minutes' notice — and the dialer honors them before fresh leads. The required disclosure is delivered and timestamped on every call.

On the roadmap — committed, not vapor

We mark unbuilt features in the product the same way we mark them here: visible, disabled, and dated.

Q3 2026

External spam-label feeds

The recordings library, DID health monitoring and the object-locked evidence vault already ship. What is dated here is the one piece they still lack: spam-label scores from the major analytics engines, which need a reputation provider we have not integrated yet.

Q4 2026

Voice-AI agents

Designed end-to-end (agent builder, live conversation monitor, AI disclosure gating) and in active development. Metered per-minute pricing — see pricing.

Want the walkthrough on your own portfolio?

Request early access and we'll import a sample of your file (or fake data, your call), run a campaign preflight, and show every screen above against real records.

Request early access