Product feature
Call data capture, outcomes and summaries from every AI call
Every Outpero call ends with a written result: an outcome label (qualified, interested, callback, not-interested and others), a short summary, sentiment, any callback time, a do-not-call flag and the custom fields you told the AI employee to capture. All of it sits next to the recording and transcript in the dashboard, in Leads & Results and in a CSV export.
* Time from the lead reaching Outpero to the first ring, under normal conditions. It depends on how quickly your lead source delivers the lead.
At a glance
Call data capture and summaries: the short version
| Details | |
|---|---|
| Runs | Automatically after every call that connects; nothing to trigger |
| You set up | The fields to capture, in Employee > Post-Call, applied the moment you save |
| You get | Outcome, summary, sentiment, captured values, callback time, do-not-call flag, recording and transcript |
| Where to read it | Dashboard, Performance, Leads & Results, All Conversations, and the call detail page |
| Recordings | The app states recordings are kept for 90 days; each one can be downloaded |
| Billing | Connected talk time only (Standard ₹5/min, Premium ₹6.8/min, plus 18% GST); each call shows its credits used |
On a call
How call data capture and summaries works, step by step
- 01The call connects, the AI employee talks to the person in their language, and the full conversation is recorded and transcribed as it happens.
- 02When the call ends, the call is classified into one outcome such as qualified, interested, callback, not-interested, missed, no_answer, busy, call_failed or voicemail.
- 03The same pass writes a short summary, judges sentiment as positive, neutral or negative from the caller's tone, and notes a callback time or an opt-out if the caller gave one.
- 04Each field you defined under Post-Call is filled from what the caller actually said, for example a budget, a preferred date or a yes or no.
- 05The result appears on the lead card in Leads & Results and as a row in All Conversations, and is sent to any post-call webhook you have switched on.
Hear how a call goes
A sample call that uses call data capture and summaries
Outbound follow-up call to a woman who enquired about a two-bedroom flat (illustrative) · Runs in Hindi with English words (shown translated); shown here in English.
- Outpero AIHello Meena ji, this is an AI assistant calling from Lakeview Homes about your enquiry for a two-bedroom flat. Is this a good time for two minutes?
- CallerYes, tell me. I am at the office, so be quick.
- Outpero AIOf course. What budget are you comfortable with for the flat?
- CallerAround seventy-five lakh, not more than eighty.
- Outpero AINoted, seventy-five to eighty lakh. And when are you planning to buy, this quarter or later?
- CallerMaybe by December, after my bonus comes.
- Outpero AIGood. Would you like our sales manager to visit or call you with the floor plans?
- CallerCall me tomorrow after five in the evening.
- Outpero AIDone, a call tomorrow at five thirty. Thank you, Meena ji, and have a good day.
- CallerThank you, bye.
Outcome: callback. Summary: wants a two-bedroom flat, budget 75 to 80 lakh, buying by December, asked for a call tomorrow after 5 pm. Captured: budget_range, purchase_timeline. Sentiment: positive. The lead moves to the callback column in Leads & Results.
An example written to show the shape of the call, not a recording or a customer result.
In the product
Set up call data capture and summaries
- 01Open Employee, choose the AI employee and go to the Post-Call tab, where the section is titled 'What the employee captures on every call'.
- 02Add a variable: give it a key such as budget_range and a label that your team will see in the dashboard.
- 03Choose its type from text, choice, boolean, number or date, and write an extract hint that tells the AI what to listen for and how to phrase the value.
- 04Save. The field applies immediately with no publish step, so the next call that is classified will fill it.
- 05Make a test call, then open All Conversations and the call detail page to check the summary, the captured values and the transcript.
In detail
Call data capture and summaries explained
What gets written after each call
A phone call is only useful to a business if someone can find out afterwards what happened. Outpero does that work for every call, so a salesperson opens a lead and reads a result, not an audio file.
| Field | What it holds | Notes |
|---|---|---|
| Outcome | One of qualified, interested, callback, not-interested, missed, no_answer, busy, call_failed, voicemail | Qualified needs a next step agreed and requirements given |
| Summary | A few lines on what was said and agreed | Also feeds caller memory when Smart follow-ups is on |
| Captured values | Your own fields, such as budget or preferred date | Typed text, choice, boolean, number or date |
| Callback time | The time the caller asked to be rung back | Used to reschedule when Smart follow-ups is on |
| Do-not-call | A flag when the caller asks not to be called again | See the Do-Not-Call controls page |
| Sentiment | Positive, neutral or negative | Judged on tone, not on the outcome |
Defining the fields you want
Fields live on the AI employee, under Post-Call. Each has a key, a label, an optional extract hint and a type. The hint matters: 'the monthly budget in lakh as a number' produces cleaner values than leaving the AI to decide.
- —Text for free answers such as the project name a lead asked about.
- —Choice for answers that fall into a small set, such as a preferred language.
- —Boolean for yes or no facts, such as 'owns a vehicle'.
- —Number for amounts and counts, and date for dates such as a planned purchase.
Where you read the results
The Dashboard shows the day at a glance and Performance shows the longer trend with a funnel and a leaderboard. Leads & Results is a kanban board where each person has notes and a history, and, if Smart follow-ups is on, a 'What the AI remembers' panel. All Conversations lists every call with filters and a CSV export.
The call detail page holds the recording player and download, the transcript, the sentiment, the summary, a talk ratio and the 'Credits used' for that call. A call-back button rings the person again from the same screen.
A summary is written by an AI from the transcript. Open the recording for any call that matters, such as a large order or a complaint, before acting on it.
Put it to work
What teams use call data capture and summaries for
| Use | How it helps |
|---|---|
| Sort a large lead list by outcome | After a bulk run, filter All Conversations by outcome so the qualified and callback leads reach salespeople first and not-interested ones are left alone. |
| Fill CRM fields without typing | Define fields such as budget, preferred date or loan amount once, and the values arrive per call in the dashboard, CSV export and webhook payload. |
| Review calls without listening to all of them | Read the summary and sentiment for every call, and open the recording only for the ones that look negative or unclear. |
| Watch a campaign's health | The Performance screen shows KPIs, a funnel, a leaderboard and a trend, so a drop in qualified calls shows up before the month ends. |
| Check what a call cost | Each call detail shows the credits used, which makes it easy to compare short and long conversations. |
Straight answers
What call data capture and summaries does not do
- !There is no per-call quality score; sentiment is a positive, neutral or negative label judged from tone, and it is not a rating of the lead or the agent.
- !A captured value exists only if the caller says it aloud, because an extract hint tells the AI what to listen for but cannot make a caller answer.
- !If the classification step fails for a call, its outcome, summary and captured values are empty, while the recording and transcript are still kept.
- !The app states recordings are kept for 90 days, so download any recording you need to keep for longer, and note that the CSV export covers the call records.
- !Qualified is a strict label: it is derived from the caller agreeing a next step and giving their requirements, so a friendly but vague call is tagged interested instead.
Call data capture and summaries: questions people ask
What call outcomes does an Outpero AI call produce?
+
Each call is given one outcome: qualified, interested, callback, not-interested, missed, no_answer, busy, call_failed or voicemail. The outcome shows in All Conversations, on the lead card in Leads & Results and in the post-call webhook payload, so you can filter, count and route leads by it without listening to calls.
Can the AI capture custom details like budget or preferred date?
+
Yes. Under Employee > Post-Call you add variables with a key, a label, an optional extract hint and a type: text, choice, boolean, number or date. They apply as soon as you save, with no publish step, and the captured values appear per call in the dashboard, the CSV export and the webhook.
Does Outpero give each call a quality score?
+
No. There is no per-call quality score. What the dashboard shows is an outcome, a summary, a positive, neutral or negative sentiment based on the caller's tone, and a talk ratio on the call detail page. To judge quality, read the summary and listen to the recording of calls that look wrong.
How long are call recordings kept?
+
The app states that recordings are kept for 90 days. Each call detail page has a recording player and a download button, so anything you need beyond that should be downloaded. The recording link sent in a post-call webhook is a signed link that expires after 24 hours, so save the file if your system needs it.
Can I export call results to a spreadsheet?
+
Yes. All Conversations has filters and a CSV export, and the Performance screen also offers an export. For results to reach another system automatically after each call, switch on a post-call webhook, which sends the outcome, summary and captured values as they are produced.
How is sentiment decided?
+
Sentiment is positive, neutral or negative and is judged on the caller's tone during the conversation, not on the outcome. A caller can be annoyed and still agree to a callback, or be pleasant and not interested, so read sentiment alongside the outcome and summary, not as a stand-in for either.
Keep reading
Other product features
Enterprise and custom: volume plans go as low as ₹3.5/min, quoted on your needs.
Get a custom quoteTry it in 30 seconds
Hear your business’s fastest employee take a call
Call the agent yourself and hear how it handles a real enquiry in your customers' language. No card, no signup — 50 free credits when you’re ready to deploy one.
No setup fee · calls from ₹5/min
Or see the full product — Calls from ₹5/min, no setup fee.