Review recorded conversations efficiently with dialogue segmented into key themes

Classify overview

Classify | Recordsure

By segmenting audio and document files into relevant themes based on natural customer-interviewer dialogue, Classify lets you quickly analyse the precise sections that are relevant to particular compliance, sales and training priorities. Many Recordsure clients use Classify to target and flag key risk factors in customer interactions, including disclosures, vulnerability, product terms, charges and attitude to risk. Automated risk analysis allows total visibility across the full breadth of recorded interactions, making spot-checks a thing of the past. Classify can be seamlessly integrated with most enterprise telephony recording systems.

What does it do?

Classify automatically breaks down conversation recordings or documents into key semantic themes to make organising  and navigating the data quick and easy.

Imagine you need to review a recording of a phone conversation. As you listen through the conversation, you categorise different segments within the conversation picking out the elements that are relevant to your review.

For instance, there might be some ‘chit chat’ throughout the conversation, there could be areas where a customer is being asked questions before advice is provided or perhaps there is a section where terms and conditions are explained and agreed to.

Similarly with text based files, you might have numerous customer touch points such as email exchanges, live chat conversations and signed documents. Organising them and segmenting manually can be a huge job, but this can be automated to perform months’ worth of work in minutes.

Recordsure uses advanced machine learning technology to automate this process. It provides a recording and transcript split into classifications which can even accommodate the complex themes found within the finance industry. The advanced speech analytics software mimics the way a human follows conversations to understand the context in which the words are used.

This is important because if you are relying on a traditional keyword search, you cannot search by underlying themes. Look at the question “what is your salary?” for instance. There are multiple different ways you could ask this using different terminology, but Recordsure is able to comprehend the meaning behind these words to interpret the deeper meaning with comparable levels of accuracy to a human.

This allows analysts to quickly organise data, efficiently target relevant areas of conversations and ignore material that is not relevant to the audit without needing to manually listen to the entire recording.

How is it used?


Efficiently review a conversation recording, honing in on the conversational topics of interest without having to conduct an end-to-end review.

Dissect conversation

Gain a complete overview of a team’s topics of conversation: how long do people spend on chit-chat? How efficient is the fact-find process? Do the best performers spend more time talking about the product or listening to the customer?

Process and complete

Quickly complete internal paperwork and documentation, jumping straight to the relevant part of each conversation and ignoring topics which are not significant.

Related technology

Triage | Recordsure


Automatically categorise conversations with unprecedented accuracy

Sentiment | Recordsure


Draw insights from conversations on the mood and sentiment of the discussion

Trend | Recordsure


Identify contextual trends in real time to keep your finger on the pulse

Assure | Recordsure


Target anomalous patterns within a conversation instead of incomplete random sampling

Digitise | Recordsure


Convert written documentation into digital formats for wider use

Transcribe | Recordsure


Speech-to-text transcription trained against both telephony and face-to-face conversations