Activity & oversight

Numbers you can open up.

Every figure in Saient leads back to the conversations behind it. See what each agent did this week, how conversations ended, where calls went, and what people keep asking, then click through to the transcript.

Get early access Book a walkthrough

7 days: of activity on every agent's card
3 outcomes: booked, handed off or answered
1 click: from a number to the transcript

A clinic owner reading a printed weekly summary over coffee

A tablet on a reception counter showing a simple bar chart

Per agent

What each agent did this week.

Every hired agent shows its last seven days: how many people it helped, the replies it sent, the visits it booked, the messages its background jobs sent (reminders, recalls, follow-ups), and how many conversations it handed to your team.

The Agents page puts all of them side by side, with which ones are working and which are still waiting on setup. Home sums it up in one line.

People, not messages.A patient who wrote five times is one person helped.
Background work counts.Reminders and recalls appear as their own number.
Handoffs are visible,so you can see when an agent leans on your team too much.

conversation.completed

"outcome": {
  "channel": "whatsapp",
  "handled_by": "ai",
  "resolution": "booked",
  "messages_in": 6,
  "messages_out": 6,
  "booking_ids": ["bkg_20"]
}

How conversations end

Booked, handed off, or answered.

When a conversation goes quiet for 30 minutes after an exchange, Saient records how it ended: who handled it (the AI, your staff or both), whether it led to a booking, a handoff or simply an answer, how many messages went each way and which bookings came out of it.

These outcomes can be sent to your CRM as conversation.completed events, with a short summary if you allow it, so your own reports see the same thing you do.

A wall calendar and a phone on a desk in late afternoon light

Calls and chat

Where calls went, and what people ask.

Voice campaigns show every call's result as it happens: answered, voicemail, no answer, retrying, and a lead score for each answered call. Studio chat agents add conversations over time, channel mix, peak hours, sentiment and the questions people ask most.

Each chat dashboard ends with suggestions drawn from those questions: the answers worth adding to your knowledge base, and the topics that keep sending people to a human.

Campaign progresscalled, answered, voicemail and retrying, live.
Peak hoursso you know when people actually reach out.
Top questionsgrouped by topic, with the conversations behind them.

Audit · one turn

find_slotsThu → 10:30, 14:00 freetool

book_appointmentConfirmed, bkg_20tool

required factsPrep note missing → rewrittencheck

webhookbooking.created → 200 OKdelivery

Audit

Every decision has a record.

Behind the numbers, Saient keeps an audit trail: each tool an agent called and what came back, each reply check that asked an agent to rewrite, each handoff and its reason, each webhook delivery and your system's response.

Usage is recorded the same way. Every agent turn is metered by model usage against your wallet, at the model tier you chose, so cost and activity always add up.

The specifics

Where to find each number.

Home: Today's visits, what needs you, and how many agents are working.
Agents page: Seven days of activity per agent, side by side.
Daily brief: Yesterday in a few lines at 7:00, for the practice manager (Ops Supervisor).
Patients: Everything one person said and did, and what each agent remembers.
Campaigns: Per-call results, lead scores and retries for voice campaigns.
Chat analytics: Volume, channels, peak hours, sentiment and top questions for Studio chat agents.
Webhook log: Every delivery attempt to your systems and its response.
Wallet: Usage and cost, line by line.

Questions

What people ask.

Can I export the data?

Yes. Send events to your CRM or a spreadsheet as they happen, or read contacts, bookings and past events through the read-only API.

Do numbers include my team's own replies?

Conversation outcomes record whether the AI, your staff or both handled them. Agent cards count the agent's own work.

How far back does it go?

Agent cards show the last seven days. Records, transcripts and events are kept, so your CRM or the API can go further back.

Is there a dashboard for clinic managers?

Home and the daily brief are built for them: what happened, what needs a person, and anything unusual.