When I describe PharmBot AI as an infrastructure company for pharmacy, I am not referring to another chatbot, another clinical application or a single piece of automation. I am describing a connected and governed foundation through which patients, pharmacists, pharmacy services and existing healthcare systems could work together more effectively. Pharmacy technology has traditionally been developed around individual transactions: one system processes prescriptions, another manages appointments, another records consultations, and separate platforms support ordering, adherence or patient communication. Although each system may solve a particular problem, the wider experience often remains fragmented. The patient moves between services without continuity, while the pharmacist must repeatedly gather information, navigate different systems and reconstruct the clinical context. The future of pharmacy will require more than digitising each of these activities separately; it will require an intelligent layer that connects them around the patient and the pharmacist.
PharmBot AI did not begin with this infrastructure vision fully formed. The original focus was prescription verification, using artificial intelligence to identify potential errors, inconsistencies and risks for a pharmacist to review. That work raised a broader question about whether AI could support the pharmacist before a prescription reached the final verification stage. This led to AIVAe, the virtual pharmacist concept, designed to interact with patients, gather relevant information and help them access appropriate pharmacy services. AIVAe then evolved into a more structured clinical tool capable of supporting Pharmacy First consultations by collecting symptoms, identifying potential red flags and organising information for pharmacist review. Each stage expanded the potential role of the technology, but it also revealed the same underlying limitation: prescription verification, consultations, triage, service booking, repeat ordering, adherence and follow-up cannot deliver their full value when they operate as disconnected products.
That realisation changed the scale of the problem PharmBot AI needed to address. Instead of continuing to build separate tools for individual pharmacy activities, the greater opportunity became the development of a modular infrastructure layer capable of supporting multiple services through a common foundation. This would allow relevant information gathered during one interaction to support the next stage of the patient journey, subject to appropriate consent, clinical governance and data protection. A structured consultation could inform a pharmacist’s assessment; prescription verification could draw attention to issues requiring professional review; and the same clinical context could support follow-up or adherence monitoring after the patient leaves the pharmacy. The infrastructure would not make clinical decisions independently or remove the pharmacist from the process. Its purpose would be to organise information, connect workflows and provide intelligent support while ensuring that the pharmacist remains the clinical authority.
The long-term vision is a pharmacy environment in which patients experience continuity rather than a series of isolated transactions. A patient might begin by describing a health concern through AIVAe, which could collect the relevant information, recognise potential warning signs and direct the patient towards the appropriate service. If a consultation is required, the pharmacist would receive structured information rather than beginning with an empty screen. If a medicine is supplied or a prescription is presented, an additional module could support the pharmacist by identifying risks or inconsistencies requiring attention. Where follow-up is clinically appropriate, the same infrastructure could help monitor progress, support adherence or reconnect the patient with the pharmacy. Each component would have a specific function, but their value would come from operating as part of one governed system with shared context and clear professional oversight.
Qatar Science & Technology Park is where I intend to begin turning this vision into an infrastructure model. The objective is not to take a finished product developed for the United Kingdom and assume that it can simply be transferred into Qatar. Healthcare systems have different workflows, priorities, regulatory requirements and relationships between patients and providers. The first stage must therefore involve understanding how pharmacy currently operates in Qatar, mapping the patient and pharmacist journey, identifying where information becomes fragmented and determining which problems would benefit most from a connected approach. Engagement with pharmacists, healthcare organisations, technology providers and other stakeholders will be essential in deciding where the first modules should be developed and validated. The technology must follow the workflow and the evidence, rather than forcing local practice to adapt around a predetermined product.
Beginning this work in Qatar also creates an opportunity to design the infrastructure from a systems perspective at an early stage. Rather than adding another application to an already fragmented environment, the aim is to understand how an intelligent layer could connect with existing clinical and pharmacy systems while remaining modular enough to evolve. The initial work may focus on a limited number of carefully selected use cases, but those modules should be designed as parts of a larger architecture from the outset. Prescription verification, AIVAe and structured Pharmacy First consultations represent the foundations of that architecture: verification provides the safety-support component, AIVAe provides the patient-facing interaction, and Pharmacy First provides the structured clinical workflow. The next phase is to connect these capabilities so that they can support a continuous pharmacy journey rather than functioning as separate demonstrations of AI.
Governance must be treated as part of this foundation rather than something applied after development. In a regulated clinical environment, every AI-supported function must have a defined purpose, clear limits and an accountable human decision-maker. Pharmacists should be able to understand what information has been collected, why an issue has been highlighted and where an AI-supported output has influenced the workflow. Activity should be appropriately recorded and auditable, with data protection, consent, access controls and clinical safety considered throughout the design. This governance-first approach is not a restriction on innovation; it is what will allow the infrastructure to move beyond experimentation and become credible within real healthcare practice.
Although Qatar will be the initial focus, the underlying challenge is international. Pharmacy systems differ between countries, but many face the same pressures: fragmented technology, expanding clinical responsibilities, growing demand, limited professional time and a lack of continuity between services. The infrastructure must therefore be capable of adapting to different regulations, formularies, languages and clinical pathways without requiring the entire platform to be rebuilt for every market. The ambition is to develop an architecture that can be configured locally while retaining a consistent foundation of governance, interoperability and human oversight.
Prescription verification was the entry point, AIVAe expanded the interaction, and Pharmacy First introduced the structured clinical workflow. Qatar now represents the opportunity to bring those elements together and begin building the layer beneath pharmacy. This is not a change in direction, but the result of understanding the problem more deeply. Pharmacy does not need more isolated technology competing for the pharmacist’s attention. It needs an intelligent and governed infrastructure that helps existing services work together, preserves clinical accountability and allows pharmacists to spend more of their time applying the judgement that technology cannot replace.