Saleslumen

The claim sits at company level and the product's own metering does not carry it.

The footer describes the company as an artificial intelligence sales platform, the questions section describes agents handling inbound and outbound calls, and the marketing page describes personalised recommendations and automated campaign adjustment. A major model provider appears among the named marketplace applications.

The pricing tiers tell a different story. Across three plans the stated differentiators are credit allowance, parallel workflow ceiling, storage volume and support level. Not one is model driven, and the vendor is otherwise precise about what each tier contains.

The credit table is the decisive evidence. It defines consumption exactly and every priced action is deterministic: standard address verification, catch all verification, resolving an address from a name and domain, and completing a workflow. If agents were consuming inference on the customer's behalf, that consumption would need a rate, and no such line exists.

The most reasonable reading is that verification, finding and workflow orchestration are the working product, and the calling agents are either newer, lightly used, or priced outside the published model.

Ask what an artificial intelligence call agent costs to run, since it does not appear in the credit table.

Autonomy runs across three surfaces and one real constraint is published as a number.

Workflows execute automated sequences between connected systems, handling lead routing and follow up without a person present, and the vendor states the parallel execution ceiling per tier at five, twenty and fifty. Publishing a concurrency limit rather than describing capacity vaguely is a genuine disclosure, because it tells a buyer how much can run at once and what the upgrade actually buys.

Campaigns run drip sequences automatically with split testing described as improving outreach on its own.

Calling agents are the highest autonomy component, handling both inbound and outbound calls, and they are the least described. Two passes located nothing on what an agent will not discuss, whether it can commit on a customer's behalf, whether it escalates to a person on uncertainty, or what a supervisor reviews afterwards.

Across all three surfaces nothing describes an approval step before a campaign starts or a message sends, and no audit trail is mentioned.

The concurrency ceiling is a capacity control rather than a safety control, so the product currently publishes limits on how much runs and nothing on what it is permitted to do.

Ask what a calling agent refuses to do and what triggers a handoff to a person.

Two passes located no model, no provider, no version, no architecture and no evaluation on any surface.

A major model provider appears in the marketplace application list, which indicates a customer can connect their own account to that provider rather than telling anyone what powers the vendor's own agents.

The metering makes the gap concrete rather than abstract. This vendor publishes its consumption rates to an unusual degree of precision, defining exactly what a verification costs, what a catch all verification costs, what an address lookup costs and what a completed workflow costs, and even stating the dollar value of a single credit. Against that standard of commercial disclosure, the complete absence of any statement about the calling agents is conspicuous: they are marketed as a headline capability and appear nowhere in the cost model.

For agents conducting telephone conversations, nothing states which speech recognition or synthesis is used, what languages are handled, what accuracy is achieved, whether call audio reaches a third party, or whether any of it trains anything.

A questions section answer on security refers to advanced encryption and continuous monitoring without naming a single specific control, and says nothing about model data handling.

Ask which provider runs the calling agents and whether call audio leaves the platform.

The testimonial section contains a checkable defect that undermines the whole set.

Six testimonials are displayed, each with a personal name, a job title and a company logo. Two of the six carry avatar images served from a public placeholder avatar service, the kind used to populate design mockups with synthetic faces before real content exists. Those two entries therefore present invented portraits as photographs of named executives. The other four use images hosted on the vendor's own domain, which does not establish they are real, only that they are not from that service.

The logo strip is separately troubling. Alongside three small named startups sit two very large consumer brands, and one of the testimonials is attributed to a manager at a global retailer describing using the campaigns module to hand influencers ready made email sequences. A manager level attribution at an organisation of that size, displayed beside placeholder portraits, is not evidence a buyer can rely on.

Two passes located no case study, no customer count, no quantified outcome from the vendor itself, and no independent review base on any major review platform.

Every quantitative claim about the product on this site sits inside a testimonial rather than being made by the vendor.

Ask for two reference customers who will take a call.

Applicable across two regulated channels and addressed on neither.

The marketing page states the purpose plainly as automating cold outreach to nurture leads through campaigns, and the platform supplies the addresses to send to through its own finder. Calling agents place outbound calls. So the vendor sources the contact, sends the email and makes the call.

Two passes across the home page, marketing page and questions section located no consent basis for contacting a resolved address, no suppression list, no unsubscribe handling or propagation, no do not call screening, no calling hour restriction and no jurisdictional guidance.

The outbound calling side is the sharper gap because it is the more heavily regulated channel and the more recently added capability. An automated agent placing commercial calls engages statutory rules on consent, calling times, identification and recording in most jurisdictions, and nothing published acknowledges that any of those exist.

The company is United States based and states domestic manufacture, so the domestic telephone consumer regime applies directly to its customers, and international customers face different regimes again.

The verification capability shows the vendor understands list hygiene as a deliverability matter; nothing indicates it treats consent as a separate question.

Ask how unsubscribes and do not call requests are captured and enforced across email and calling.

Terms and a privacy policy are published and linked, and a dedicated security page exists in the footer, which was not opened during this review and is recorded as existing rather than assessed.

What can be assessed is the questions section answer on data security, and it is boilerplate. It refers to advanced encryption techniques, compliance with data protection regulations, and continuous monitoring and updating of security measures, without naming a single algorithm, framework, regulation, certification or control. Every clause in it could be written by any vendor about any product.

Two passes located no data processing agreement, no subprocessor list, no retention period, no processing location and no described data subject request route.

The holdings are substantial for a platform at this price. Customer relationship records, complete email campaign histories, resolved contact addresses for individuals who never dealt with the vendor, support case content, and audio or transcripts from calls handled by the agents.

That last category is the one requiring the most careful handling and receives none. A platform conducting telephone conversations on a customer's behalf holds recordings of third parties, and nothing states whether they are retained, for how long, or who can hear them.

Ask for the processing agreement, the retention period on call recordings, and the subprocessor list.

The vendor supplies contact data as a metered product and discloses nothing about where it comes from.

The credit table prices finding an email address from just a name and a domain at ten credits, charged only when an address is found. That is a resolution service producing personal data about an individual, and it is a core priced capability rather than an incidental feature.

Two passes located no source, no provider chain, no aggregator, no licence basis, no consent basis, no coverage statement by geography, no accuracy figure and no refresh cadence.

One commercial term genuinely favours the buyer and is credited. Charging only when an address is found, and charging nothing when a verification returns unknown or error, aligns the vendor's revenue with delivered results rather than with query volume. That is the correct incentive structure and it is the opposite of the standard credit model, where a failed lookup still bills.

Good incentives on billing are not the same as provenance, and the person whose address is resolved and then contacted has no relationship with anyone here.

A peer built earlier in this session names eight upstream suppliers on its own pricing page, so this disclosure is achievable in the category rather than commercially impossible.

Ask where resolved addresses originate and on what lawful basis they are held.

The extraction exposure common in this index is absent. Two passes located no scraping utility, no professional network automation, no profile harvesting and no browser extension.

Integration runs through sanctioned commercial interfaces. The marketplace names applications from both major productivity suites, a major model provider, a large data vendor, a general automation platform, two major customer relationship systems and a messaging platform. Each of those is a published interface with its own terms, accessed the intended way.

The scripting model deserves a note. Customers can build private internal applications, discover public ones in the marketplace, or publish their own for other customers to install. A publishing model means third party code runs inside other customers' accounts, which raises a review and vetting question the vendor does not address: nothing states what approval an app passes before publication, what permissions it can request, or what happens if a published app misbehaves.

The email finder raises a provenance question rather than a platform terms one, and is graded there.

Calling runs through telephony infrastructure that is not named, so whose carrier terms bind the customer is unstated.

Ask what review a published marketplace app undergoes and what data it can reach.

Three artefacts exist and none carries substance that could be verified in this review.

A security page and a status page are both linked from the footer, and a developer portal runs on its own subdomain. Having all three at this size is more structural intent than many peers show.

The developer portal is the finding. Its page title is the unmodified default generated by the web framework used to scaffold it, which means the portal was deployed without the placeholder metadata being replaced. That is a small detail and a telling one for an axis measuring operational care, and it echoes the same pattern found on another record in this session where advertised developer surfaces did not resolve.

On substance, two passes located no encryption specification, no access control description, no tenant isolation statement, no retention period, no incident history, no responsible disclosure route and no training commitment.

The stewardship exposure is wider than the price suggests. The platform holds relationship records, campaign histories, support cases, resolved third party contact data, and whatever the calling agents capture from telephone conversations. Whether any of that trains a model is unstated.

A status page on the main marketing domain rather than an independent service is worth noting, since it shares fate with the site it reports on.

Ask whether customer data or call content trains any model.

Squarely applicable through the calling agents and unaddressed.

The questions section states that the platform provides agents for both inbound and outbound calls. A person telephoning a customer of this vendor, or receiving a call from one, is therefore speaking to software. Two passes located no disclosure position of any kind: no statement that an agent identifies itself, no configurable disclosure line, no default opening script, no recording notification, and no guidance to customers on whether their jurisdiction requires any of these.

Recording consent is the specific unaddressed duty. Many jurisdictions require both parties to agree before a call is recorded, and an automated agent that produces transcripts is recording by construction.

On the email side the position is better. Campaigns send from the customer's own connected accounts under their own identity, so the sender is real even where the sending is automated.

One further authenticity question arises from the vendor's own marketing rather than its product. A site presenting invented portraits as photographs of named executives, established on the evidence axis, is itself a disclosure failure of a kind, and a buyer weighing how this vendor handles representation of identity may reasonably factor that in.

Ask whether a calling agent can be configured to identify itself and announce recording.

Reasonable breadth for a vendor this size, with a structural ambition that outruns its documentation.

The marketplace names eight popular applications spanning both major productivity suites, two major customer relationship systems, a large data provider, a general automation platform, a messaging platform and a model provider. A documented interface and webhooks are offered for custom work. An internal scripting model lets customers build private applications, and a publishing route lets them release apps for others, which is a genuine platform ambition rather than a connector list.

The qualifications matter. The developer portal carries the default page title left by the framework used to build it, which suggests it was published before it was finished. Two passes located no interface reference with endpoints, no rate limits and no authentication documentation.

One integration claim rests on weak ground. A model context protocol server is referenced only inside a customer testimonial, and that testimonial is one of the two whose avatar is a placeholder image. No vendor surface reviewed here documents such a server, so its existence is not established.

Storage is published per tier at three, twenty and fifty gigabytes shared, which is a real published limit.

Ask whether the interface documentation is complete and whether a model context protocol server actually exists.

Two passes located no hosting provider, no region, no country of processing, no residency option, no single tenant arrangement, no customer managed keys and no subprocessor list.

The only geographic signals are a footer stating domestic manufacture and a company base in Miami. Neither is a statement about where customer data rests, and a company's headquarters and its infrastructure are routinely in different jurisdictions.

The absence matters more than the price point suggests because of what the platform accumulates. Relationship records, campaign histories including message content, support cases, resolved contact data on individuals who never dealt with the vendor, and audio or transcripts from calls handled by the agents. Call content in particular is the category where processing location determines which authorities can compel access and which recording rules apply.

The questions section answer on security refers to compliance with data protection regulations without naming one, which for a buyer outside the United States is not an answer to the transfer question at all.

The marketplace connects to third party providers whose own processing locations are equally undisclosed, and no subprocessor list exists to identify them.

Ask in which country customer records and call recordings are stored and processed, and whether any residency commitment is contractual.

A dedicated security page exists and is linked from the footer. It was not opened during this review and is recorded as existing rather than assessed, so a buyer should read it directly rather than treating this grade as a judgement on its contents.

What two passes did establish across the home page and questions section is that no certification is claimed anywhere a prospective buyer would encounter it. No attestation is named, no auditor, no assessment period, no framework, no penetration test, no completed questionnaire, no subprocessor list and no responsible disclosure route.

The security answer in the questions section is the vendor's most prominent statement on the subject and it names nothing: advanced encryption techniques, compliance with data protection regulations, and continuous monitoring. Under the credential test there is no verb attaching any credential to this company, and no scope boundary, because no credential is asserted at all.

A status page exists, though it runs on the main marketing domain rather than an independent service, so it would likely fail alongside whatever it reports on.

The practical position is that a buyer with any procurement requirement has nothing to submit from the surfaces most buyers actually read.

Ask what the security page contains, whether any attestation exists or is in progress, and for a disclosure contact.

The strongest axis on this record and genuinely well executed.

Three tiers publish an exact per user monthly figure alongside every quantitative limit that matters: credit allowance at thirty, sixty and ninety thousand a month, parallel workflow ceilings at five, twenty and fifty, shared storage at three, twenty and fifty gigabytes, and the support level attached to each. A thirty day trial applies to all paid plans with no card required.

The credit table is the standout and it does something almost nothing else in this index does. It defines each priced action exactly: standard verification costs one credit when the result is conclusive, catch all verification costs ten, resolving an address from a name and domain costs ten and only when an address is found, inclusive of any verification, and a completed workflow costs twelve. Inconclusive and error results are free.

Then it states the value of a credit at one tenth of a cent, which lets a buyer convert any action directly into dollars and compare against any competitor. Publishing the unit value is rare and it is the thing that makes the rest usable.

One commercial term is unusual and favourable: agencies may add client accounts on any plan at no charge.

Against this, no annual option or discount is published, and a section higher on the same page renders with empty placeholder labels.

Ask whether annual billing is available.

Two structural factors help and the specifics are undocumented.

Only monthly per user rates are published with no annual commitment offered, so there is no term lock: a customer who wants to leave is not waiting out a year. That is the most consequential portability property a subscription can have and it follows from the pricing design rather than from a promise.

A documented interface and webhooks exist at every tier, giving a technically capable customer a route to extract records programmatically without depending on an export feature, and the internal scripting model provides a further path.

What two passes could not locate is any export route, file format, statement of export contents, termination provision, notice period or post cancellation retention and deletion position.

What accumulates spans several categories: relationship records, campaign structures and their performance history, support case threads, resolved contact data, and whatever the calling agents captured. Storage is published per tier, so the volume is bounded and the retrievability is not described.

The unfinished state of the developer portal is relevant here, since the interface is the primary exit route and its documentation could not be confirmed as complete.

Ask what an export contains, whether call recordings are included, and what is deleted on cancellation.

One control is done properly and the rest of the discipline is absent.

Address verification is built into the product rather than sold beside it, and the handling is more careful than most. Standard verification returns valid, invalid or catch all for a single credit. Catch all domains, which are the hardest case and the commonest source of unexpected bounces, get their own dedicated ten credit verification that resolves them to valid or invalid rather than leaving them ambiguous. Inconclusive and error results cost nothing, so a customer is not billed for a verification that told them nothing.

Treating catch all resolution as a distinct priced capability shows the vendor understands where bounces actually come from.

Everything else is missing. Two passes located no mailbox warm up, no domain health monitoring, no sender authentication guidance, no inbox placement testing, no spam or content checking, no sending throttle, no sender rotation and no published delivery or bounce figure from the vendor.

Two testimonials claim bounces fell to near zero through verification. Those sit in the same set that includes entries with placeholder portraits, so they carry no evidential weight and are not credited.

Verification is the highest leverage single control, so the product covers the most important thing and nothing around it.

Ask what warm up and authentication support exists, and what bounce rate customers actually see.

Positioning is broad to the point of being uninformative, with two genuine signals underneath it.

The questions section states the platform suits businesses of all sizes from startup to established enterprise, and the marketing page claims fit for business to business lead generation, business to consumer engagement and commerce conversion simultaneously. Those are three different motions with different buyers, and claiming all of them tells a buyer nothing about whether it fits them.

Two real signals do emerge from the commercial design. Per user pricing between thirty and ninety dollars with no seat minimum and no annual commitment describes a small team or an individual operator rather than an enterprise deployment. And agencies are addressed as a deliberate segment through a term that lets them add client accounts on any plan at no charge, reinforced by a managed service partner initiative and an affiliate programme.

The agency angle is the most coherent part of the positioning and the site does not lead with it.

Geographic and language coverage is unstated. The company is Miami based, the footer states domestic manufacture, the site is published in English only, and two passes located no supported country list, no language statement and no data coverage claim for the address finder by territory.

Ask what address coverage looks like outside the United States.