AI call quality assurance · by aQb Solutions

Automated Call Quality Assurance with Evidence-Linked Scores.

CallLense is AI call quality assurance software that scores every call against your scorecard, with a quote from the call behind every answer.

Video transcript

The FCA's Consumer Duty requires firms to put their customers' needs first. CallLense by aQb Solutions turns call recordings into scored audits, with a quote from the call behind every answer. Calls are scored against question sets, with weighted categories and pass thresholds that admins can edit. Pull a call from a HubSpot ticket or any other source, or upload a recording by hand. Each call is transcribed automatically, with every speaker identified. Name the analysis, choose a question set, and the AI scores every question and validates the evidence. The report gives the score, a summary, strengths, weaknesses and recommendations. Every answer is backed by a quote from the call, and anything below the pass threshold is flagged. The score can sync back to the HubSpot ticket, automatically or on demand. Score every call. Show the evidence. Coach with confidence.

CallLense by aQb Solutions is AI call quality assurance software that transcribes every call, scores it against your scorecard and backs each score with a timestamped quote from the call. It turns recordings into scored audits. Pull calls from HubSpot or any other source, or upload them by hand, and each one is transcribed with its speakers identified. Choose a scorecard, and the AI scores every question, backs each answer with a timestamped quote from the call, and flags anything below the pass threshold, so managers can review and coach with evidence.

Impact you can measure

Illustrative targets, to be confirmed in a pilot on your own data.

100%Calls scored, not a spot check
MinutesTime from call to scored report
Hours → MinutesTarget time to review a call
100%Answers backed by a quote from the call
SecondsTo find the moment behind any answer
Same dayFeedback to the employee
100%Reports with strengths, weaknesses and recommendations
24/7Call analysis

What it does

Call ingestion from any source

Pull calls from HubSpot or any connected source, or upload recordings by hand, and file them by library and customer.

AI call transcription

Each call is transcribed automatically with speakers identified and timestamps, and any line jumps to that moment in the audio.

Question-set scoring

Score every call against weighted question categories with pass thresholds. Question sets are editable, and questions that do not apply are excluded.

Evidence-backed answers

Every answer quotes the moment in the call, with its time, and the quote is checked against the transcript before it is shown.

Scored reports and review flags

Each report gives the score, summary, sentiment, action items, strengths, weaknesses and recommendations, and flags anything below its pass threshold for a manager to review. Flags assist review and are not proof of compliance.

Coaching and performance tracking

Assign each analysis to an employee and track benchmarking and progress over time, with every run kept in History.

Key Questions Answered

01 / 04

Which recording sources and languages are supported?

CallLense (sometimes searched as CallLens) takes calls from HubSpot or any connected source, or are uploaded by hand. Transcription and speaker separation depend on audio quality, and language coverage is confirmed on your own recordings before you rely on it.

How it works

  1. 01Pull or Upload
  2. 02Transcribe
  3. 03Choose Question Set
  4. 04Analyse & Score
  5. 05Validate Evidence
  6. 06Review Report

Technology stack

  • Next.js
  • NestJS
  • Python
  • PostgreSQL
  • AWS: S3 & SQS
  • AI model transcription
  • AI-based scoring

Use cases

01Regulated financial services

Evidence behind every customer call

Score every call against your conduct and disclosure questions, with the exact quote behind each answer. Scores support review, not proof of compliance.

Want to see CallLense on your own calls?

Book a working session

Last updated