Skip to main content

Automation overview

Insighto.ai gives you three automation layers that work together to turn a conversational assistant into an end-to-end operations engine. You decide how a single conversation branches, what happens before and after it, and how bookings get distributed across your team — all from the dashboard, with no engineering involved.

Most production deployments use all three layers in tandem. A caller hits a workflow at the front door, the assistant runs a conversation flow during the call, and a calendar pool picks the right teammate when an appointment gets booked.

The three layers at a glance​

LayerScopeBest for
Conversation FlowInside one conversationStructured intake, branching on extracted answers, deterministic paths the assistant must follow
WorkflowsAcross conversations and over timeAfter-hours routing, reminders 24 hours before a booking, missed-call follow-up
Calendar PoolsBooking distributionRound-robin across SDRs, priority for senior reps, sharing slots across treatment rooms

How they fit together​

Where to use each layer​

Use Conversation Flow when the assistant must collect specific information in a specific order, branch on what the customer said, or call a tool at a precise moment in the conversation. Examples: a dental clinic that must capture insurance details before offering appointment slots; a fintech assistant that runs KYC checks before any account-related question can be answered.

Use Workflows when something has to happen before, after, or instead of the conversation. Examples: a real estate brokerage that forwards out-of-hours calls to an on-call agent; an e-commerce store that texts customers 24 hours before a curbside-pickup appointment; a restaurant that follows up on missed calls with an SMS containing a booking link.

Use Calendar Pools when more than one teammate handles bookings and you need a fair (or deliberately unfair) distribution. Examples: a SaaS sales team round-robining demos across AEs; an enterprise practice routing VIP appointments to the senior dentist first; a multi-location service company assigning jobs to whichever technician is free.

Four real deployments​

Dental practice — Bright Smile Dental, Austin. Conversation Flow drives a new-patient intake (name, DOB, insurance, reason for visit). When the caller confirms an appointment, the assistant books on a Calendar Pool that distributes across three chairs by least-busy. A reminder workflow texts the patient 24 hours before, and a missed-booking workflow asks them to reschedule.

Real estate brokerage — coastal property firm. An inbound-call workflow checks the time. Inside business hours it routes the caller to the assistant; after hours it forwards directly to the on-call agent's mobile. New seller leads that go through the assistant trigger a conversation.closed event that pushes the lead into HubSpot.

E-commerce — vegan meal-kit subscription. A chat assistant handles delivery-window changes and refund questions. When a customer requests a callback, a workflow places an outbound call from the on-shift CX rep's number. Calendar Pools aren't needed; conversation flow and workflows are.

B2B SaaS — analytics platform. A sales SDR assistant qualifies inbound leads via a conversation flow (industry, team size, current tool). Qualified leads book a demo through a Calendar Pool that round-robins across AEs. A follow-up workflow sends the booking confirmation, then nudges them 24 hours and 1 hour before the call.

Where to start​

If you're new, build a working assistant first with a system prompt and one data source. Once that handles 80% of your traffic well, layer in:

  1. Conversation Flow — to enforce a structured intake step that the assistant otherwise drifts from.
  2. Workflows — when you spot a recurring "something happened, now we need to react" pattern in your operations.
  3. Calendar Pools — the moment two or more people are taking bookings.

Where to next​