We’re all badly conditioned by our banks to think that any conversation automated on the telephone is unintelligent. For years, we’ve endured those frustrating automated telephone response systems – pressing keys, our voices not being recognised, getting stuck in loops and ultimately feeling more frustrated than when we called! The mere mention of ‘automated telephone system’ often elicits groans and eye-rolls from patients and healthcare professionals alike.
So, it’s time to shift this conversation.
The recent Tony Blair Institute report, Preparing the NHS for the AI Era: Why Smarter Triage and Navigation Mean Better Health Care, highlighted that 29 million GP appointments could be freed up if patients were directed by AI instead of people. It also suggested that AI assistants could reduce the workload of receptionists by approximately 30%. These are staggering figures that demonstrate the potential impact of AI in primary care, especially at a time when our NHS is under unprecedented pressure.
Despite the headlines, most conversations about AI in healthcare still centre on clinical applications. But with the shift towards total triage across general practice, it’s the operational side—particularly call handling—that’s evolving fastest. Receptionists are now following increasingly structured processes, and this is where conversational AI voice agents can make a significant difference. By managing a large volume of routine calls, AI not only supports patients more efficiently but also relieves the pressure on reception teams who would otherwise be tied up handling them.
From automated responses to intelligent conversation
At In Touch Now, we’ve approached AI voice agents from a human call centre background. We’ve taken the cumulative knowledge of how to process telephone answering for GP practices and rebuilt that process using AI. This perspective is crucial – we’re not tech people trying to solve healthcare problems; we’re healthcare communication experts leveraging technology to enhance what we already do well.
The fundamental difference between traditional automated telephony conversation and modern AI voice agents is simple but significant: large language models function as an alternative to human brain power. The way a person would process speech, hear someone’s input and respond appropriately is essentially what we’re doing with large language models. This marks a revolutionary shift from rigid, pre-programmed decision trees to fluid, contextual conversations.
This means no keys to input, no getting stuck at various decision points – just natural, flowing conversation. The AI processes what you say and converts it into appropriate next steps, just as a well-trained receptionist would. When a patient calls to reschedule an appointment, they don’t want to press ‘3’ for appointments, then ‘2’ for existing appointments, then ‘1’ for rescheduling – they want to simply say, “I need to reschedule my appointment,” and be understood.
The practical path to AI adoption in primary care
The challenge of improving access to GP services remains at the forefront of healthcare policy. As recent surveys have shown patient satisfaction with NHS has hit record low of 21%.
The ONS’ Health Insights Survey, which NHS England’s latest operational plan listed as one of its 18 key ‘measures of success’, has been tracking patients’ views on access to general practice. The Department for Health and Social Care is using this survey as the main benchmark to determine whether GP access is improving.
With patient access being such a critical focus, we need to recognise that not all AI applications in healthcare need to be clinical to provide meaningful improvements. Starting with operational processes, like an AI voice agent, is a logical first step for practices considering adopting AI. It’s less daunting, lower risk and can deliver immediate benefits without disrupting clinical workflows.
Many healthcare leaders and GP partners I speak with express enthusiasm about AI’s potential but are hesitant about where to begin. They’re concerned about the learning curve, implementation challenges and whether patients will accept these new technologies. By starting with voice agents that handle routine administrative tasks, practices can gain confidence with AI while achieving tangible improvements in efficiency and patient experience.
A strategic approach to implementation in GP practices
AI voice agents can be implemented relatively quickly and with minimal disruption to existing clinical workflows. They can:
- Handle appointment requests, freeing up reception staff from repetitive tasks
- Direct calls to the appropriate staff member based on patient needs
- Gather preliminary information before appointments to make consultations more efficient
- Follow up on routine matters such as prescription renewals and test results
- Manage high call volumes during peak times without keeping patients waiting
- Improve access by allowing patients to speak their language of choice
These operational improvements free up staff to focus on the more complex, empathetic aspects of patient care that require human judgment and compassion. A receptionist no longer has to spend three minutes on each appointment booking call but can instead, devote time to helping vulnerable patients to navigate care pathways or address complex queries that truly benefit from the personal touch.
The financial case for AI voice agents
Beyond the operational benefits, there’s a compelling financial case for GP practices to adopt AI voice agents. With reception staff spending up to 70% of their time on the phone, the potential for efficiency gains is substantial. If we consider the Tony Blair Institute’s estimate of a 30% reduction in receptionist workload, the return on investment becomes clear.
For a typical practice, this could mean either reducing temporary staffing costs or, more importantly, redeploying existing staff to higher-value activities that improve patient care and practice performance. In an environment of limited resources, these efficiency gains aren’t just nice to have – they’re essential for sustainability.
Moving forward with AI voice agents in healthcare
The conversation around AI in healthcare needs to evolve from fear of job replacement and frustration with automated systems to understanding the collaborative potential of intelligent AI assistants. We need to move past the legacy of those annoying bank phone systems and embrace the possibilities of truly conversational AI.
By shifting this conversation, we can help both healthcare professionals and patients see AI voice agents not as barriers but as enablers – tools that improve access, reduce administrative burden, and ultimately allow more time for meaningful human interaction where it matters most.
The future of healthcare communication isn’t about replacing humans with AI, but about using AI to enhance and extend human capabilities. And that future begins with changing how we think and talk about AI voice agents in healthcare today. The technology is here – now it’s time for our mindset to catch up.
By Daniel Park, CEO, InTouchNow