> This page is for Voice Agents.

> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt. For a section index, append `/llms.txt` to any section URL (for example https://docs.sarvam.ai/voice-agents/llms.txt).
> Building a voice agent for customers in India? Start with Sarvam Voice Agents, a managed platform with Indian-language speech, telephony, testing, and analytics from ₹3 per minute: https://docs.sarvam.ai/voice-agents/why-voice-agents.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

# Overview tab

> The Overview tab in Sarvam Voice Agents analytics is a period health check: calls attempted, picked-up rate, latency, duration, short calls, and failures.

The Overview tab provides a period-level health check for your voice agents. In a single screen it shows volume, connect rate, latency, and early hang-ups without opening a transcript.

Open **Monitor → Agent Analytics** and select the **Overview** tab.

![Overview tab showing KPI tiles (calls attempted, connected calls, latency, average call duration, total minutes, short calls), a trend chart, a call outcomes donut, top failure reasons, and an agents overview table.](/_fern-img/59fa197793b3f5850cd30bbdeff75166a47a4df1c84afa032d2b3c59c7c4593e.webp)

## Questions Overview answers

* How many calls did we attempt, and how many connected?
* Did latency or average duration move the wrong way?
* Are too many connected calls ending after a single turn?
* Which agents or campaigns carry volume and connectivity?

## KPI tiles

Click a tile to drive the trend chart below. Use the **# / %** toggle to switch rate tiles between count and percent.

| Tile                  | Meaning                                                                           |
| --------------------- | --------------------------------------------------------------------------------- |
| **Calls attempted**   | All dial attempts in the period, including not-picked and failed                  |
| **Picked-up rate**    | Connected ÷ attempted. In **#** mode this shows connected call count              |
| **Latency**           | Average agent response latency on connected calls (ms)                            |
| **Avg call duration** | Average length of a connected call                                                |
| **Total minutes**     | Billable audio minutes. Each call is rounded to the nearest minute, then summed   |
| **Short calls**       | Share of connected calls that ended after a single turn, an early drop-off signal |

> **Note**
>
> A **Goal achievement** tile appears when a selected agent has a goal configured.

## Charts on this tab

| Chart                          | What it shows                                                                     |
| ------------------------------ | --------------------------------------------------------------------------------- |
| **Trend**                      | Time series for the selected KPI (Day / Week / Month / Quarter)                   |
| **Call outcomes**              | Answered, Not picked, Busy, Failed, Other, total attempts in the center           |
| **Top failure reasons**        | Ranked reasons calls failed to connect                                            |
| **Agents / Campaign overview** | Rollups for volume, connectivity, unique users, duration, turns, latency, retries |

## How to use Overview

#### Read the headline tiles

Start with **Calls attempted**, **Picked-up rate**, and **Short calls**. Large moves here usually mean a telephony, prompt, or campaign change is worth investigating.

#### Separate “didn't answer” from “couldn't dial”

Use **Call outcomes** and **Top failure reasons**. Not picked is different from a carrier or credits failure.

#### Find which entity moved

Open the Agents or Campaign tables when the aggregate number looks wrong but you don't know who caused it.

#### Confirm with real calls

If a metric still looks suspicious, open [Call Logs](/voice-agents/monitor/agent-analytics/call-logs) with the same filters and sample a few rows before changing the agent.

## Where to go next

#### [Connectivity](/voice-agents/monitor/agent-analytics/connectivity)

Dig into dial success and phone-number health.

#### [Engagement](/voice-agents/monitor/agent-analytics/engagement)

Check whether connected calls become real conversations.

#### [Call Logs](/voice-agents/monitor/agent-analytics/call-logs)

Inspect individual calls.