Call analytics
Call analytics and post-call analysis for AI voice agents
When a call ends, AI reads it and records a summary, a disposition, the caller’s sentiment and the fields you ask for. Each call’s data lands in call history, a 12-month dashboard, your webhook and the API.
How it works
From a finished call to numbers you can use.
Post-call analysis runs on every call your agents take. You choose what it captures; the rest is automatic.
In the agent’s Post-call summary settings, add the fields you want from each call: a name, what the AI should look for, and the answer format (Text, Yes / No or Selector). Put them in the order you want.
When the call ends, AI reads it and records a short summary, a disposition, the caller’s sentiment and your fields. The caller is never asked these questions: the answers come from the finished call.
Each call’s page keeps its summary, transcript, recording and post-call analysis together (unless HIPAA mode is on), so the full record of a call is one click from the list.
The Analytics dashboard adds up your calls by month, for any of the last 12 months, and exports them. The same analysis goes to your post-call webhook and is available through the platform API.
In action
The call analytics dashboard
Disposition and sentiment for a month of calls.
Extraction
The fields you need, typed and in order.
Tell the AI what to pull from each call in plain words. Each field has an answer format, so the data arrives ready to count and route.
Text, Yes / No or Selector
Pick how each answer comes back: written text, a simple yes or no, or one of the options you set.
Answers from your own list
A Selector field can only come back as one of its options, so “Reschedule” never arrives as “moving my booking”.
In the order you set
Drag fields, or use the arrows, to put them in the order you want them listed.
Summary, disposition and sentiment built in
Every call also gets a short summary, a disposition and the caller’s sentiment, without adding a field.
Dashboard
A month of calls on one screen.
The Analytics dashboard adds up every call your agents took, for any of the last 12 months.
The headline numbers
Total calls, the share completed, average duration and the share that reached voicemail, with volume by hour of day and your top agents.
Dispositions and sentiment
How calls ended and how callers felt, as counts and percentages of the month.
Your fields, counted
Each extracted field shows how often it was captured and its most common answers; Yes / No fields show the share of each.
Export
Export the month’s calls from the dashboard to work with them in your own tools.
Data out
Every call’s data, where you need it.
Each call’s analysis is saved with the call and sent on, so your own systems get the same fields you see.
Post-call webhook
Fires when the call ends, with the full summary and analysis. Failed deliveries are retried, and each call shows its delivery status.
Platform API
Call history, with each call’s summary, analysis and recording, is available through the platform API for your own dashboards.
Workflows when a call finishes
Start a workflow when a call finishes and act on its fields: branch on an answer, send an email or SMS, or write the outcome to a connected system or a Google Sheet.
Bookings from the calendar’s answer
Booking outcomes in call history and webhooks come from the calendar system’s result, not from what the agent said on the call.
Frequently asked questions
What is post-call analysis?
When a call ends, AI reads the conversation and fills in a short summary, a disposition (how the call ended), the caller’s sentiment and any fields you have defined. The agent never asks these questions out loud: the answers are worked out from the finished call.
What data can the AI extract from a call?
Any field you define in plain words, such as the reason for the call, whether an appointment was booked, or whether the caller qualified. Each field is Text, Yes / No or Selector; a Selector answer is always one of the options you set. Fields are listed in the order you choose.
Which voice AI KPIs can we track?
The Analytics dashboard shows total calls, the share completed, average call duration, the share that reached voicemail, volume by hour of day, top agents, dispositions and sentiment, for any of the last 12 months. For first-call resolution or lead qualification, add a Yes / No field such as issue_resolved or lead_qualified: the dashboard shows the share of Yes for the month. Measure the same things on your current line first, so you have a baseline to compare against.
How do we measure call deflection ROI without hurting customer satisfaction?
Use dispositions to see how many calls the agent completed and how many it transferred, sentiment to watch how callers felt, and a Yes / No field to track whether their issue was resolved. Export the month’s calls, or pull them through the API, to set them against your own costs. Sentiment is the AI’s reading of the call, not a survey score.
How does call sentiment analysis work?
After each call, the caller’s sentiment across the whole conversation is classed as Positive, Neutral, Negative or Mixed (Mixed when it changes significantly during the call). The dashboard shows the month’s split, and each call’s value is in its post-call analysis.
Does every call get an AI summary?
Yes. Every call gets a short summary of why the caller rang, who they were if they said, and the outcome. It is on the call’s page in call history and in the post-call webhook.
Can we send call data to our CRM or BI tool?
Yes. The post-call webhook sends each call’s summary and analysis when it ends, with retries and a delivery status on the call. Call history is also available through the platform API, a workflow can write outcomes to connected systems or a Google Sheet, and the dashboard exports the month’s calls.
Are booked appointments taken from what the agent said?
No. Booking outcomes in call history and webhooks come from the calendar system’s result, so a call where the agent said “you’re booked” but the booking failed isn’t reported as booked.
How is this different from Call Review?
Call analytics records data from every call automatically and adds it up on a dashboard. Call Review is weekly quality assurance: AI reads a sample of calls against your own pass criteria, reports what went wrong and suggests fixes.
See what your calls are telling you.
Book a demo and we’ll set up post-call fields for your calls and walk you through the dashboard.