GP Patient Access and AI Reception Insights | QuantumLoopAI

AI Reception for PCNs: A Practical Rollout Guide

Written by Alina Zhudro | November 7, 2026

AI reception at PCN scale should make patient access easier across several practices while keeping each surgery’s responsibilities clear. Start with the patient journey, prove where requests arrive and measure the work left for staff before expanding.

A shared purchase can look straightforward on paper. In practice, one surgery may use a different consultation tool, another may have a branch site, and a third may serve patients who need more help on the phone. A useful rollout plan makes those differences visible.

At QuantumLoopAI, EMMA’s focus is AI patient access for NHS GP surgeries during surgery hours. This guide sets out a practical way for PCN leaders, practice managers and ICB teams to assess that role across their own services.

What should a PCN agree before choosing AI reception?

Agree what a successful patient contact should achieve. Answering the telephone is one step. The next steps are capturing the request, making it available to the appropriate team and giving the patient a clear understanding of what happens next.

NHS England’s telephone journeys guidance, last updated in May 2024, describes PCN working through common technology, common processes and shared call handling. Its emphasis on understanding each practice’s starting point remains useful when assessing an additional patient access service. Check current local requirements separately from that document’s historical funding and programme dates.

Create a short practice map before the supplier demonstration. Include opening hours, telephone provider, consultation tool, request destinations, human support arrangements and who owns each stage. A common contract does not mean every practice has an identical workflow.

Follow a request all the way to the practice team

A persuasive voice demonstration is only the beginning of a useful assessment. Ask to see what happens after the call ends, using the product and configuration proposed for your practices.

StageWhat the PCN should establishEvidence to request
Patient contactWhen AI reception operates and how a patient gets human helpA demonstration of the agreed access journey
Request captureWhich information is collected and how corrections are handledAn example using fictional patient details
Practice workflowWhere the request appears and who picks it upA walkthrough in the relevant practice configuration
Staff actionWhat staff still need to review, enter or completeA clear account of the remaining work
Patient next stepWhat the patient has been told to expectThe wording and responsibility agreed with the practice

Be clear about patient outcomes. Ask which routine appointment tasks each practice will offer, what the patient hears when an action is completed and which requests still need a response from the team. Agree the next step for patients when an option is unavailable.

EMMA by QuantumLoopAI supports patient requests and routine appointment booking and cancellation during surgery hours. The options available depend on the service agreed with each practice.

EMMA Hub gives teams visibility of submissions and business intelligence. That makes the demonstration a chance to review both the patient conversation and how the practice monitors the work that follows.

Keep human support visible throughout the journey

EMMA does not provide clinical advice or clinical triage. The rollout should make the boundary between patient access and clinical decision-making clear to staff and patients.

Consider a fictional example: a patient begins with an administrative request, then says they need to speak to someone. The assessment should establish how the patient reaches the practice team and what information the team receives. It should also establish what happens if that team is temporarily unavailable.

Agree how patients can ask for human help, how staff review the interaction and how concerns are investigated. Include reception staff and patient representatives in that review. A smooth demonstration for one confident caller cannot establish accessibility for everyone.

Build an assurance pack that matches the deployment

NHS England’s digital clinical safety guidance, updated in March 2025, distinguishes the manufacturer’s DCB0129 responsibilities from deployment and use under DCB0160. It explains that ICB support does not remove a practice’s responsibility for operating a digital solution safely.

Ask your clinical safety and information governance leads which evidence applies to the proposed deployment. Keep the product version, intended use and local responsibilities attached to the review. A supplier logo or assurance badge gives less useful information than the underlying document and its scope.

A working evidence index can record the document owner, version, review date and unresolved questions. It should cover the applicable safety documentation, data processing arrangements, accessibility assessment, service support and business continuity plan. Keep confidential assurance documents in the appropriate controlled process.

How QuantumLoopAI approaches the service

QuantumLoopAI’s current internal evaluation, signed off by its Clinical Safety Officer, reports 99.6% red-flag detection. This is an internal evaluation result. It should be considered alongside the proposed workflow and local assurance, rather than treated as a guarantee for every live call.

The current service supports 19 languages signed off through QuantumLoopAI’s clinical safety review process. It also offers website chat and disconnected-call follow-up. Those capabilities give a PCN specific patient journeys to discuss during evaluation.

For procurement, EMMA has an official G-Cloud 15 listing under QuantumLoop Technologies Ltd. QuantumLoopAI also confirms participation in the RM6200 Artificial Intelligence DPS. Ask the buyer’s procurement team to confirm the appropriate route and current requirements. RM6200 uses a further competition; supplier participation does not remove that step.

Compare the whole service cost and the work it saves

A low headline price is useful only when the PCN understands what it buys. Ask every supplier for the same view of the proposed service: participating practices, operating hours, included usage, implementation, support, additional charges, renewal and exit arrangements.

For a usage-based offer, examine the unit being charged and any allowance or cap. For a fixed offer, establish what is included and which changes affect the price. A fair comparison keeps those conditions visible.

Then look at staff work. A captured request may still require review and action. Measure that work rather than treating every answered call as a completed patient need. The business case should show how reception capacity can be used, alongside the cash costs of the service.

QuantumLoopAI includes number porting and the line as part of its supplied service, bringing the surgery’s existing number into that service with one provider to contact. Review any existing telephone contract and agree the transition plan. If another proposed arrangement relies on forwarding calls, ask the current provider whether forwarding attracts charges under your contract.

For an EMMA proposal, ask QuantumLoopAI to relate the quotation and expected workflow to your practices’ needs. A complete explanation is more useful than a saving percentage without a baseline.

Use a staged rollout with a clear expansion decision

The following is a suggested planning sequence, rather than a promise about implementation time.

  1. Define the starting point. Record demand, access difficulties, staff work and the practice differences that affect the proposed service.
  2. Agree the operating model. Assign owners for requests, human support, patient communication, assurance and supplier issues.
  3. Review the proposed configuration. Walk through the relevant systems with fictional patient details. Record what is live and what remains unresolved.
  4. Begin with an agreed scope. Choose participating practices and a review period that allow the PCN to learn without assuming one site represents every patient population.
  5. Decide whether to expand. Review patient experience, staff work, outstanding requests and service issues against the starting point.

Set the expansion criteria before the launch. For example, agree how the PCN will investigate a request arriving in the wrong place, who can pause a rollout and what evidence is needed before another practice joins. The decision then rests on an agreed process.

What evidence should PCN and ICB leaders review?

Use measures with clear definitions. Call-answering time, time to human support, requests reaching the intended workflow and staff time spent on follow-up describe different parts of the service. Report them separately.

Keep each practice visible within the PCN results. An overall average can hide a problem at one site or in one patient group. Review patient feedback and staff observations alongside the counts.

Published customer experience can add context. In QuantumLoopAI’s Conisbrough Group Practice case study, Practice Manager Bex Cottey says, “Don’t take EMMA off us.” That is a specific practice’s experience. Ask which parts of the deployment are relevant to your own service and what evidence you would want to see locally.

Questions about AI reception for PCNs

Does a PCN need identical systems at every practice?

Start by mapping the systems and workflows already in use. Establish the supported configuration for each practice before agreeing a rollout. A shared procurement decision should make the differences manageable and visible.

Does answering a call mean the patient’s request is resolved?

No. Call answering, request capture, staff action and resolution are different events. Agree the definitions before comparing supplier results or reporting benefits.

Can a successful practice deployment prove an ICB-wide result?

A practice deployment provides evidence about that setting. A larger decision should examine the additional sites, workflows, support needs and patient populations it would cover.

Where should a PCN start with QuantumLoopAI?

Arrange an EMMA demonstration with your practice map and priority access problems. Ask QuantumLoopAI to walk through the proposed patient journey, identify the supported configuration and explain the evidence needed for a PCN decision.