About Dr Rajarshi Mitra

A practising surgeon bringing clinical judgement to medical AI.

I evaluate whether AI-generated medical answers are clinically accurate, properly reasoned and safe to act on. I also build practical digital systems for doctors and clinics—drawing on problems I first had to solve in my own practice.

Explore my Medical AI work View the case study
  • Practising laparoscopic surgeon
  • Medical AI evaluation
  • Clinician-in-the-loop workflows
  • Digital systems for doctors

MBBS · MS (Surgery) · FACS · FICS · FIAGES

Diploma in Laparoscopic Surgery · Strasbourg, France

Portrait of Dr Rajarshi Mitra
Why These Are Connected

Clinical practice taught me to look beyond the confident answer.

For more than 20 years, I have made decisions in situations where omissions, misplaced confidence and delayed escalation can matter. That experience shapes how I examine medical AI: not only whether an answer sounds fluent, but whether its reasoning is clinically defensible, its uncertainty is handled honestly and its advice is safe for the situation described.

My work with digital systems began from a different problem. I found that many generic marketing approaches did not understand how a medical practice actually works. So I began building and testing practical systems around visibility, patient information, review workflows and ongoing maintenance.

These are two distinct areas of work, but they share the same discipline: define the real problem, inspect the evidence, preserve human judgement and build a system that can be reviewed and improved.

Two Areas of Work

Clinical evaluation for Medical AI. Practical systems for medical practices.

Medical AI evaluation and clinical governance

I work at the point where medical language becomes a clinical decision. The focus is whether an AI-generated answer is accurate, properly reasoned, appropriately cautious and safe enough to proceed to human review.

  • Evaluate medical answers for clinical accuracy and evidence adherence
  • Assess clinical reasoning and differential diagnosis
  • Develop difficult medical cases and evaluation rubrics
  • Identify unsafe advice, unsupported claims and missing escalation
  • Design clinician-in-the-loop review and release controls
  • Review AI-assisted medical content and patient communication
Explore Medical AI →

Digital systems for doctors and clinics

I help doctors examine and strengthen the systems around their online presence. The work is grounded in first-hand experience of building and maintaining these systems for a real medical practice—not in generic agency templates.

  • Google Business Profile structure and maintenance
  • Local visibility and competitor analysis
  • Ethical review-generation workflows
  • Website and enquiry-pathway review
  • Patient-information content systems
  • Ongoing monitoring and improvement
Explore Clinic Growth →
Selected Work

A clinician-governed medical content system

The featured case study documents how a medical-content workflow was simplified from a multi-stage pipeline to one principal article-generation step without removing its clinical release controls.

Evidence collection, deterministic technical checks, clinician review, bounded correction, explicit publication authorization and recorded live verification remained distinct stages. Retained project artifacts document clinician-directed corrections involving laparoscopic port placement, biliary anatomy after gallbladder removal and pain-location laterality.

My contribution centred on clinical boundaries, finished-content and medical-illustration review, bounded correction, and material I selected or authorized for publication. Repository integration, deployment and live-route verification remained separate technical stages.

The workflow became simpler. The clinical responsibility did not.

Read the full case study See all case studies
How I Work

Four principles behind the work

Clinical consequence before surface polish

A fluent answer can still be unsafe. I look first at reasoning, omissions, uncertainty and escalation—not merely grammar or presentation.

Evidence before certainty

Claims should remain within the boundaries of the available evidence. When the evidence is incomplete, the answer should say so.

Human judgement at release

Automated checks can support a workflow. They do not replace the clinician responsible for deciding whether medical material is ready to proceed.

Inspectable systems

A useful system should leave enough structure and retained evidence for its decisions, corrections and limitations to be examined.

Clinical Background

The experience behind the evaluation

I am a practising laparoscopic surgeon in Abu Dhabi with more than 20 years of clinical experience. My qualifications include MBBS, MS (Surgery), FACS, FICS, FIAGES and a Diploma in Laparoscopic Surgery from Strasbourg, France.

Clinical experience does not make every medical AI judgement automatically correct. It provides the context needed to recognise when an answer is medically incomplete, when its reasoning does not support its conclusion and when escalation or human review is required.

The Medical AI and digital-systems work presented on this website is independent, carried out through 1Health LLC, and is not offered on behalf of any hospital or healthcare institution.

I provide clinical evaluation and governance support. I do not present this work as regulatory or legal advice.

Work With Me

Bring me the work that needs clinical judgement.

If you are building a healthcare AI product, evaluation programme or medical-content workflow, tell me where clinical review is needed. If you are a doctor or clinic working on visibility and digital infrastructure, use the clinic-growth pathway instead.

Discuss a Medical AI project Explore Clinic Growth
Connect on LinkedIn