# Metricsense

> Metricsense reads 100% of a company's customer conversations (calls, reviews, tickets and surveys), surfaces any insight you define in plain English, and links every finding to the exact source quote. Built and operated by Avesta Labs.

Metricsense analyses every conversation, scores both the customer side and the agent side, and deploys inside the customer's own sovereign cloud with PII stripped before any transcript reaches the model. No tagging and no model training: you write what to find in plain English and it runs. One engine serves two buyers:

- **Call analytics** for teams that sell and serve by phone (insurance and other regulated contact centres): catch the compliance slip, the misstated fact and the deal lost to a missed follow-up in the calls nobody reviews. Both sides scored, every flag quote-linked, inside your own sovereign cloud. Proven with a free pilot on around 50 of your own calls.
- **Feedback intelligence** for product and CX teams: find the crash buried in 5-star reviews and the theme hiding across thousands of tickets. Ask any question and get the exact quote behind the answer. Free to start.

The flagship reference customer is Life Insurance Direct, an Australian life-insurance broker. They run their own calls through Metricsense inside their own sovereign cloud: 10,000+ calls read and scored on both sides, twenty-five insurance-specific concepts built from their own documented compliance process and validated by them, and no new QA analysts hired. A flag is a candidate issue with the transcript moment attached, for the licensee's own compliance team to review, not a finding and not a breach.

## Core pages

- [Home](https://metricsense.ai): any conversation, any metric. Product overview and a live "ask a question" demo.
- [Call analytics for insurance](https://metricsense.ai/insurance): read 100% of adviser calls, flag ASIC/AFCA compliance risk the same day, every finding quote-linked, inside your own sovereign cloud.
- [Data governance](https://metricsense.ai/governance): where an AFS licensee's call recordings sit, what crosses a boundary, the instruments that apply, and the independent assurance behind it.
- [Reviewing every call versus sampling](https://metricsense.ai/vs/manual-call-qa-sampling): why manual QA sampling is a selection problem rather than a volume problem, and where sampling still wins.
- [Feedback intelligence](https://metricsense.ai/feedback): read 100% of reviews, tickets and surveys and answer any question with the exact quote behind it. Free to start.
- [FAQ](https://metricsense.ai/faq)
- [Pricing](https://metricsense.ai/feedback#pricing)

## Key capabilities

- **Analyse 100% of conversations**, scoring both the customer side and the agent side.
- **Custom metrics**: define any insight in a plain-English prompt. No tagging, no model training.
- **Evidence-linking**: every score traces to the exact transcript moment or the exact customer quote.
- **Per-agent Call Quality Index (0 to 100)**, recurring-issue ranking, and threshold alerts.
- **Deploys in your own sovereign cloud and region**; PII is stripped before any transcript reaches the model.
- **Multilingual**: validated on your own languages during the pilot.

## Use case pages

- [Call transcript analysis](https://metricsense.ai/use-cases/call-transcript-analysis)

## Comparisons

- [Metricsense vs Kraftful](https://metricsense.ai/vs/kraftful)

## About and contact

- Built and operated by [Avesta Labs](https://avestalabs.ai/).
- Contact: hello@metricsense.ai
- [Privacy policy](https://metricsense.ai/privacy-policy)
- [Platform login](https://platform.metricsense.ai)