Coffee with Callbi – New Features You Need to See

Introduction: A Milestone Session

Our eighth edition of Coffee with Callbi was one of the most engaging yet. We welcomed a full house of CX professionals, QA leaders, and contact centre managers to hear about the latest developments in Callbi’s speech and text analytics platform. This session marked a significant milestone in our journey. Not only have we surpassed the 200-customer mark across South Africa’s BPO and contact centre landscape, but we also unveiled a series of features that elevate Callbi from an analytics platform to a comprehensive quality and insight ecosystem.

Rating Sheets: Closing the QA Loop

The highlight of the session was the introduction of Rating Sheets, a feature that connects objective and subjective quality assessments. Traditionally, QA teams have run two parallel processes – automated scoring through queries and manual evaluations in spreadsheets. Rating Sheets finally unify these processes within Callbi.

Objective measures are automatically populated through queries, often covering 80 percent of a scorecard, leaving only the softer skills such as empathy, product knowledge, and tone for manual review. Built-in workflow means each sheet can be assigned, actioned, commented on, and tracked with a full audit trail. This ensures nothing slips through the cracks. By freeing up assessor time, QA teams can focus on coaching and mentoring agents rather than spending hours ticking boxes.

Clients piloting the feature have reported significant time savings, with some evaluations being 60 to 70 percent faster, while also enhancing agent engagement and closing the feedback loop within a single platform.

The Callbi Assistant: Your Built-In Knowledge Partner

We also introduced Callbi Assistant, our own hosted large language model designed to make quality teams more productive. Think of it as your resident call centre expert built directly into the Callbi interface.

Ask how to create a rating sheet or what silent time means, and you receive a clear, actionable response. The knowledge base is continuously updated with Callbi documentation, best practices, and FAQs. It does not learn from client data, remaining aligned with our commitment to responsible AI. This serves as the first line of support for busy QA teams, reducing support requests and accelerating the adoption of new features.

Executive Dashboard: Insight at a Glance

The updated Executive Dashboard offers a clear overview of performance trends. Operational leaders and executives can now determine if interventions are enhancing efficiency, supported by ROI data showing an average reduction of six to ten per cent in handle time.

Language distribution insights help in planning staffing for multilingual contact centres, ensuring the correct mix of agents is available. Negative sentiment drill-downs allow teams to pinpoint where conversations went wrong and respond to root causes quickly.

Transcript Improvement: Making the Model Smarter

A powerful feature demonstrated in the session is the ability to teach Callbi. If a transcript misinterprets a product name, acronym, or unique phrase, you can highlight the error, submit the correction, and let the platform learn from it. These small adjustments are combined to improve the accuracy of the models for all clients.

Only a tiny snippet of audio, usually five to ten seconds, is shared for model training, never an entire call, and never with personally identifiable information unless explicitly authorised.

Responsible AI and Security

Security and privacy remain central to Callbi’s approach. The platform is ISO 27001 certified and hosted on AWS in Ireland (GDPR compliant) and South Africa (POPIA compliant). Annual penetration testing ensures a strong security stance. Access control is entirely managed by the client, and even our support staff cannot access your environment without explicit authorisation.

We do not use client data to train our models. Instead, we depend on curated knowledge bases, public domain information, and user-submitted training snippets under signed data-sharing agreements.

Preparing Teams for Change

Change management was a lively topic of discussion. Some organisations worry that automation might replace QA roles. In reality, Callbi elevates the role of the QA team. The focus shifts from manual checking to generating insight and providing business intelligence to operational and executive management.

Clients who have successfully adopted Callbi report that their QAs are now positioned as Insight Analysts, adding strategic value and driving measurable improvements in performance and customer experience.

Live Q&A Highlights

Participants asked about telephony integrations and real-time coaching. Callbi integrates with almost any telephony platform that allows call extraction, including 1Stream, Smartz, Vicidial, Telviva, QContact and others. Although Callbi is mainly a post-call analytics tool, an Agent Assist feature is currently being tested. This will offer real-time compliance prompts during live calls and is expected to be launched within the next three to six months.

Conclusion and Next Steps

This session showed just how far Callbi has come, from powerful speech analytics to a complete quality management and coaching solution.

If you missed the live demo or want a personalised, one-on-one walkthrough of these new features, we would be delighted to arrange a session tailored to your environment.