All FAQs

AI-Native Operations

How much of the platform an AI agent can stand up, run, and watch on your behalf.

17 questions answered

Can the platform be set up and managed by AI instead of an ops team?

Yes. Every operation — offers, buyers, contracts, campaigns, caps, dispositions, reports — is available through a fully documented REST API, and a built-in AI assistant can operate the entire platform conversationally. Your whole build can be stood up and changed by an AI agent instead of days of manual clicking.

Related: AI Suite

Is there an AI assistant inside the platform?

Yes. A built-in AI assistant handles conversational setup and analysis in the admin. Say "set up a new offer" or paste a partner’s posting specs, and it extracts the fields, asks only about the gaps, and executes the configuration.

Related: AI Suite

Can the AI assistant actually create buyers, contracts, and campaigns, or does it only answer questions?

It takes real actions. The assistant sits in a panel on every admin page, reads your live data, and can operate the entire platform — creating and editing offers, buyers, contracts, campaigns and caps — after interviewing you for the details it is missing.

Related: AI Suite

Does the AI ever change anything without asking me first?

Writes are held behind an explicit approval step: the assistant shows you exactly what it is about to do, and a configuration change never runs until you approve it. Chat transcripts are stored with personal data redacted, so there is a reviewable trail of what the AI did.

Can the assistant tell me why a buyer stopped receiving leads?

Yes. It ships with nine packaged setup and diagnostic procedures — including routing diagnostics, lead investigation, and contract and balance checks — so you can ask why a buyer went quiet and get back the specific gate that is blocking them.

Can I talk to the assistant instead of typing?

Yes — voice in and voice out are both built in. The assistant can also navigate the admin for you, highlight the exact control it is describing, and run a guided walkthrough step by step.

Related: AI Suite

What happens when the assistant cannot finish what I asked?

You can escalate to a human from the same panel — a support ticket on standard plans, live in-panel chat on priority support — so you never lose the context or start over in a separate helpdesk.

Can I manage the platform from Claude Code or another MCP client?

Yes. We publish an open-source MCP server on public npm, so you can query core objects — leads, buyers, contracts, campaigns, offers, people, reports — and manage buyer and contract status, from Claude Code or any MCP client with a scoped API key. The admin includes a setup tab with copy-paste configuration.

Related: AI Suite

Do I need to bring my own AI API key, and how is AI usage billed?

No key needed — the AI runs on our platform credentials. Usage is metered against your credit balance at actual cost plus a small markup and itemized by session and by day, so you pay for what you actually run rather than a flat AI tier.

Can AI be switched off for my account?

Yes. AI is a per-account switch, so a tenant that would rather not use it can turn it off entirely and keep running the platform through the UI and API.

Can AI set up a new buyer from their posting spec?

Yes. Paste or point us at a buyer’s posting documentation and it is parsed into a ready delivery action — posting URL, authentication, field mapping, body template, and the response rules that decide accept, reject or duplicate.

Related: AI Suite

Does the platform monitor itself?

Automated anomaly detection watches routing and volume patterns and runs automated root-cause analysis — surfacing problems like a buyer silently rejecting or a source drying up before a human would notice them in a report.

How quickly would you spot a buyer that quietly started rejecting leads?

Anomaly scans run every 15 minutes across acceptance rate, delivery failures, volume, margin and invalid-field rate, measured against the same hour on your own prior days in your own timezone. Critical alerts arrive with a diagnosed likely cause, not just a number that moved.

Can I version a lead scoring model and test it before it goes live?

Yes. Scoring models are versioned with an explicit draft-to-live promotion step, and versions can be validated, exported and imported — so a scoring change is proven before it ever touches routing.

Can the platform forecast revenue or suggest where to change price and caps?

Both, automatically. A seven-day revenue forecast with upper and lower bounds is projected from your own history, and an optimization view ranks concrete suggestions — an under-margin contract to review, an under-used cap you can adjust in one click.

Can the platform call leads automatically after delivery?

Yes, through an optional AI voice dialer billed on usage. It works a configurable multi-attempt cadence, enforces 8am–9pm calling windows in the consumer’s own timezone, honors do-not-call lists and on-call opt-out requests, and transfers a qualified consumer live to the buyer.

Related: AI Suite

What does AI-native actually save?

Setup that takes days on legacy platforms becomes a conversation, and ongoing changes — new buyers, new sources, rule changes, reporting questions — are handled in minutes. That cuts both the implementation cost and the ongoing headcount needed to run the system.

Still have a question?

Pricing is a flat plan plus published usage rates, and a build is configuration rather than custom development. Start where it makes sense for you.