About Dr Ryhan Hussain

Learning to build useful healthcare AI.

Online
Medical education creator
2014 – 2021
London
Healthcare data, then NHS doctor
2022 – 2025
San Francisco
First employee, cancer AI startup
2024 – 2025
London
Venture capital · Part-time healthtech due diligence advisor
2024 – 2025
Doha
Chief of Operations, clinical AI
2025 – now

How I got here.

I’m Dr Ryhan Hussain, a former NHS doctor based in Doha, Qatar. I now work in healthcare AI product development, partnerships and operations, with experience across the UK, the US and Gulf health systems. My role often means helping patients, clinicians and engineers understand each other’s needs and turn them into a workable product.

I qualified in medicine with an MBBS in 2022, after also completing a BSc in management in 2020. Working across more than ten medical and surgical specialities taught me to make decisions with incomplete information, ask for help, and think carefully about what an error could mean for a patient. Those habits still shape how I approach AI.

My first projects were smaller. I made educational videos for medical-school applicants and later built a tool that moved doctors’ shifts from a spreadsheet into their calendars. Colleagues shared it, but a wider rollout proved harder. It was an early lesson that building something useful and making it fit an organisation are different jobs.

Since then, I’ve joined a cancer AI startup as its first employee, advised a venture fund on healthcare technology, and moved into clinical AI operations in the Gulf. I’ve learned most by working closely with the people affected by a product, especially when their feedback changes what we thought we should build.

Selected work

What that work has involved.

Patient research and cancer AI

At an early-stage cancer AI startup, I built a patient advocacy board of more than 75 members and ran over 50 structured interviews with patients, carers and oncologists. Their feedback helped shape a six-month product roadmap.

One useful direction was helping patients prepare questions for their oncologist. I helped develop personalised appointment questions and integrate clinical-trial matching. The aim was to help people take part in conversations about their care, while keeping treatment decisions with them and their clinical team.

Clinical AI reliability and evaluation

For a cancer second-opinion tool, I helped define how it would find relevant clinical evidence and cite its sources. This approach, called retrieval-augmented generation (RAG), was paired with clinical test cases, human review and a fallback when the evidence was insufficient.

When an oncologist flagged an unsafe summary, I reproduced the error and traced it to how the system handled cancer staging. We added a check and a test case to catch it in future versions. I also personally reviewed reports at launch. These checks helped us find problems; they were never a reason to assume every answer was correct.

Hospital AI pilots in the Gulf

My work with health systems in Qatar and the UAE has included choosing a useful first workflow, preparing clinician training, and working with engineering teams on hosting patient data within the required country. I have also led information-security and regulatory preparation, including work to secure ISO/IEC 27001 certification.

In one UAE pilot, clinicians asked us to focus on a quality-improvement audit tool instead of the wider set of features we had proposed. We narrowed the work around that need. The pilot was still being evaluated in July 2026, so I treat it as a delivery and learning example, rather than evidence of a proven clinical outcome.

Healthtech strategy and team building

Alongside product work, I helped an early-stage team grow from one to eight people through structured recruitment. I also worked on which patient needs to prioritise and how a cancer second-opinion product might create value for patients and those paying for their care.

In a separate, part-time venture advisory role, I reviewed the clinical and technical claims of healthcare startups. That meant checking whether the evidence supported the proposed use, how a product might fit into care, and which assumptions still needed testing.

How I work

Lessons I keep coming back to.

Start by listening

Clinical experience gives me a starting point, but it does not tell me what every patient or team needs. Interviews, observation and follow-up help test my assumptions. Some of the most useful product decisions have come from changing my mind after those conversations.

Test the ordinary working day

A tool needs to fit around appointments, handovers, existing records and limited time. I try to understand those constraints early, then watch how people use it. Training, access and follow-up deserve as much attention as the feature itself.

Be clear about the limits

An answer that sounds convincing can still be wrong. I care about traceable sources, meaningful tests and knowing when a person should review the result. I have also declined features that would give patients prognosis estimates beyond what the product could safely support.

Writing and teaching

Sharing what I’m learning.

I started with free medical-school guidance that reached four million video views. Today I share plain-English explanations of healthcare AI and lessons from moving into product work.

Explore the AI lessons, read the clinical career playbook, or book a conversation about your own work.

Watch on TikTok