The call dock
A draggable dock that is the dialer, and stays with you as you move around the app. Keypad, minimize and consent-gated recording, never a modal that traps the screen.
Roadmap
No dates, no teasers. Each item says what it is and where it stands. Items move between columns as they ship.
Launch path
Now (current stage)
Evrment is built and used every day by its founding agency.
Next
A small group of agencies works in Evrment and shapes what ships.
Then
The waitlist is invited in first, with founding-user benefits.
Launch
Evrment opens to every agency, with published pricing.
37 items
A draggable dock that is the dialer, and stays with you as you move around the app. Keypad, minimize and consent-gated recording, never a modal that traps the screen.
Real calls from the browser, and real inbound calls routed by your business hours and whether you are clocked in. When you are not available, callers reach an honest voicemail.
Works a list three dials deep and three laps around, tagging outcomes as it goes and resuming where you stopped. Every lead keeps a permanent dial history.
Your script sits at the center of the call with the lead's details filled in. A linter warns on false authority, invented urgency and affiliation claims, with a citation, and never blocks your own words.
Built from twelve real lead-vendor file formats, with automatic column mapping and saved per-vendor mappings. Duplicates are reported instead of silently dropped, and every batch is kept in history.
Your internal do-not-call list is enforced on every outbound path, and imports are screened against it before a number is dialable. Suppression survives lead deletion and is tracked by phone and by email.
Automation is held to 9am to 8pm in the lead's own time zone, Monday to Saturday. A human reaching out after hours gets a confirmation step and an audit entry, not a block.
Recording is off by default and gated for two-party consent states. Turning it on is a deliberate choice, and the rule applies to inbound calls too.
Every action is recorded with an actor, a module, an action, a subject and a before-and-after diff. Filter it, export it to CSV, or open any lead's full activity history.
A fifteen-tag taxonomy with editable SMS, voice and email copy at every step, mapped to where each person is in their lifecycle. Customer care, referral, birthday and policy-review automations are built in.
Pipeline, book of business, commissions, share ledgers, P&L and month close in one place. Promotion and recognition tracking runs off the same numbers.
A channel-based team hub for your downline, and booking pages backed by consent records. Agent microsites and vCards come with every seat.
Carrier logins, licenses and tools in one place, masked by default. Every reveal is recorded in the audit log.
Real sign-in, organizations and roles are live for private-access users. Self-serve accounts open with early access.
Every lead keeps the batch it arrived in and the vendor it came from, so a source can always be traced to what it produced. Nothing is re-keyed from one screen to another.
Every lead carries where it came from: its source, its UTM source, medium and campaign, and the lead group it arrived in. The attribution starts at the moment of capture instead of being reconstructed later.
Stages move on real events — a disposition, an appointment, an approved policy — instead of on someone remembering to drag a card. The pipeline reports, it does not ask to be maintained.
Calls, texts and email against the same person sit in one thread, with the phone hub, texts and email as views over it rather than separate tools.
Campaign inbound and your personal line are answered by different rules. Routing is live; conversational AI voice for Ava and Rea is in development and labelled that way.
What changed, why it matters and who should act, brought to you instead of waiting in a dashboard. The Monday brief runs on the same numbers. AI-written wording is simulated until model keys are wired.
"Call me in three months" becomes a dated wake-up on the record, and anything the system should not decide alone is flagged for a person. The organization remembers, not one producer.
Every kind of event is set to interrupt you, wait for the digest, or stay silent. Attention is a budget, and you set it.
A five-rung ladder from observing to executing approved work, with an approvals queue, role permissions, escalation and an audit entry behind each step. Nothing runs above the rung you set.
The work clock and the activity tracker keep human effort and system effort from blending into one number, so a leader can see what people actually did.
Scripts, rebuttals, carrier notes, comp structures and training are the agency's editable content. Evrment ships no one else's material. AI filing of new content is simulated until model keys are wired.
A coach that follows the stage of the call and prompts the next question. It runs on a deterministic simulation today, not a live model, and the product says so wherever it appears.
Policies, the servicing queue and the post-sale record live beside the lead they came from. The sale is the start of the record.
What stays on the books, tracked against the same policy record as the sale, with the customer-care schedule, reviews, birthdays and referral asks running off it. Carrier data feeds are a separate, unbuilt item.
Lead spend tied to the production it created, expenses, the share ledger and the annual picture, with a freshness stamp on every number. Evrment does not move money.
Recruit, interview, contracting, licensing by state, onboarding, a seat, a front door and a place on the tree — one record the whole way, instead of a spreadsheet and a folder.
Channels, threads, reactions and locked rooms inside the system that runs the business, so a win posts itself from approved business and a conversation sits beside the record it is about.
A lead posted from a form or a vendor lands as a record with its source and campaign attached, is checked for a duplicate on the formatted phone number, and enrolls on the spot if it matches an audience.
Book, reschedule and mark an appointment showed or missed, against the same record that holds every call, text and email. A reschedule keeps the history instead of starting over.
Booking, a missed appointment, a sale and a chargeback each move the pipeline and enroll the right follow-up by themselves, with an audit entry for the move.
Reminder steps counted back from the appointment rather than forward from the booking, on text and on voice, with the name, number, date and time merged in.
Marking an appointment missed enrolls the recovery campaign within the hour, after the do-not-call list and the audience rules have been checked.
A booking page that shows times in the visitor's own time zone and stores a consent record with the appointment, versioned with the wording they agreed to.
12 items
A portal for your policyholders: coverage, documents, beneficiaries, messages and service requests, branded to your agency.
Ava and Rea answering and placing calls in natural conversation, inbound and outbound. They will identify themselves as AI and hand off to you when a human is needed.
Campaign and one-to-one email sent from your own domain, with proper authentication records. Replies land in the same conversation as calls and texts.
Self-serve plans, seats and invoices inside the app. Add a seat for a new recruit without talking to anyone.
Your booking page on a public link that prospects can open from a text or an ad. Consent is captured on the page and stored with the appointment.
Register your numbers, monitor how carriers label them, and show your name on the caller's screen. A healthy number is answered more often.
Rich messages with your verified brand, buttons and cards on phones that support them. Texts remain the fallback everywhere else.
Content, social, funnels, forms, ads and trend research inside the operating environment. The surfaces are built; most connect to outside services as you wire them, and say so before you click.
Field-level protection for the most sensitive records an agency touches. The requirements are written; the implementation is not shipped, and no page claims otherwise.
Every appointment can carry a briefing that shows on the calendar: where they came from, what they asked and why they agreed to meet. Today it is written by whoever booked it. Writing it from the conversation itself is in development.
Enrollments already carry the time their next touch is due, and automation advances when the app is opened. Sending on schedule with nobody in the app needs messaging registration and a background runner.
Ask about the business in your own words and get an answer from your own records, with the records it read shown beside the answer. Runs on the deterministic engine today; model-backed reasoning is in development.
41 items
Policy status pulled from carriers instead of typed in. Your book would update itself when an application moves.
Each agency in a hierarchy gets its own branded workspace, rolling up to the one above it. Built for IMOs and large downlines.
Work your list and take inbound calls from your phone. The same queue, script and dispositions as the desktop.
A manager can listen in and coach during a live call, with the agent's consent. Training that happens on real calls instead of role play.
Flags pressure language in real time, while there is still a chance to change course. The spoken version of the script linter.
Notes and follow-up reminders written from the call transcript. You review them instead of typing them.
When a deal is EFT-approved, a win post goes out to your team and, if you choose, your social accounts. Underwriting alone never counts as a sale.
Flowstate learning from your own recorded patterns, where you win and where calls stall. Advice specific to you, not a generic playbook.
A single gate every number passes before it can be dialed, combining DNC, quiet hours, consent and litigator screening. One answer, with the reason attached.
Reach clients on the channel many of them already use every day. Same consent rules, same conversation history.
Connect Evrment to the rest of your stack. Leads, dispositions and policy events available to your own tools.
Single sign-on and automated user provisioning for larger organizations. Seats follow your directory.
Practise the objection you keep losing — price, the spouse, existing coverage, coverage through work, "I need to think about it" — before the next live call.
Training that answers what the calls actually showed, assigned to the producer it applies to, with the change measured on the calls after it.
Connecting an agency's own ad accounts so performance can be read into Evrment and leads can arrive with their campaign and creative attached. Meta is the first planned connection. Nothing is connected today.
Daily and campaign caps, a department budget, approval thresholds, a maximum change size, automatic pause rules, escalation, rollback and an audit entry behind every change. No AI unit changes spend above an approved bound.
Sending what actually happened after the click — contacted, booked, showed, applied, issued, still on the books — back to a connected ad platform where the platform supports it, so delivery can learn from business rather than from form fills.
Source to campaign to creative to lead to outcome, reported as known, assisted and unattributed, with a confidence on each. Attribution is never presented as certainty.
Continuous monitoring of the agency's connected accounts, platform and policy changes, public ad libraries where permitted, and search and seasonal trends. Every report says what changed, why it matters and what to do.
Flags a spend spike, a broken tracking pixel, delivery that stopped, a cost-per-lead jump, a landing-page conversion drop, duplicates, a form failure, pacing drift, creative fatigue, a rise in opt-outs, or downstream quality falling away.
A hypothesis, then variants — copy, static concepts, video concepts, hooks, scripts and shot lists — then compliance review, then approval, then a test, then what was learned kept on the record. Drafting is what Vega does today, on a simulation.
Landing pages, forms, qualification, thank-you and booking flows, with variants and conversion tests, under the agency's own brand, offer, audience and rules. Agent sites, digital cards and booking pages with consent records already run today.
Shield checking an ad, a form or a page against the agency's own standards, carrier advertising requirements, platform policies and consent and disclosure rules, and flagging it for a human. The script linter is the live ancestor of this.
Weighing speed-to-lead, unworked leads, calendar availability, show rate and close capacity before recommending more spend — so more leads are only recommended when the floor can absorb them.
Surfacing untouched leads, no-contacts, missed callbacks, no-shows, stale opportunities, clients due a review, referral moments and inactive recruits as an alternative to new spend. Subject to consent, opt-out, quiet hours and the internal do-not-call list.
Routing a new lead by licensed state, product authorization, language, availability, capacity, ownership and team, with the rules visible and auditable. Routing never uses protected characteristics.
The objections and questions leads actually raise on calls, reported as aggregate themes for marketing to answer. Never a client's own words or details.
The ad, the promise, the offer and the questions it tends to create, on the record before the producer picks up the phone.
Comparing vendors and campaigns on contact, set, show, placement and persistency instead of on cost per lead — including how often a vendor sells you someone you already own.
Recruiting ads and funnels, licensed and unlicensed paths, interview booking, and the measurement that matters: source to activated producer, not source to application.
Hypotheses, tests, results and decisions kept as the agency's own history, so the lesson stays when a partner or an employee leaves.
Giving an outside agency or consultant a scoped seat with approved performance context, while the attribution, the history and the memory stay with the agency.
A short morning read from Atlas and Vega: what moved, what needs a decision, and what to leave alone. It sits beside the Monday briefing that already runs.
Ask the system what the agency already knows across its own records and playbooks, instead of asking the founder the same question again.
A confirmation page that introduces the person they are meeting, what the conversation covers and how long it takes, with one tap to confirm, reschedule or cancel.
The qualifying questions from your own fact-find asked and captured in the conversation, on your script, and written to the record as answers rather than as notes.
Stepping into a thread the AI is holding, mid-conversation, with the handover and everything said before it written to the audit log.
A leader reviewing what was actually said, with a summary beside the recording, inside the recording-consent rules that already apply.
Received, contacted, in conversation, set, confirmed, showed and recovered, broken out by lead source and by campaign, so a cheap lead that never shows is seen for what it costs.
Atlas handing a question to the unit that owns it, collecting what comes back, and reporting once, instead of you asking each one.
Atlas answering outside the app, by text message and by phone call, with the same scope and the same audit trail it has inside it.
Changelog
Join the waitlist and hear first when the next stage opens.
Want to see it early? Request a private preview