Applied AI and retrieval
Retrieval-augmented workflows, extraction, classification and LLM-assisted systems with evaluation and human review designed in.
Founder expertise
Dr Musaab Elkheder · Founder and Principal Consultant
A clinically trained software engineer, health data scientist and delivery leader who combines domain understanding with hands-on system design, evaluation and implementation.
Building software since 2007, with experience spanning research systems, health information, mission-led organisations, multi-tenant platforms and complex programme delivery.
01 · Capability
The useful work happens where these disciplines meet around a real operational problem.
Retrieval-augmented workflows, extraction, classification and LLM-assisted systems with evaluation and human review designed in.
Reproducible analysis, statistical modelling and data-quality assessment that can be inspected rather than accepted on presentation alone.
Django, multi-tenant architecture, PostgreSQL, APIs and data pipelines built for testing, documentation and handover.
Clinical training and health-data experience applied to records, terminology, interoperability, coding support and operational workflows.
Scope, prioritisation, delivery risk, stakeholder alignment and handover across complex programmes and technical work.
English/Arabic products and right-to-left interfaces designed as part of the workflow, not added as a cosmetic translation layer.
02 · Experience and domain record
Selected professional contexts show the breadth behind the RefLens offer.
Charitable sector transformation: led digital transformation for a London-based charitable trust, including data platforms, reporting and AI-assisted internal workflows.
Foundation-scale programmes: directed strategic programmes and digital transformation for a Doha-based foundation.
Co-founded software ventures and delivered hospital, university-library and clinical reporting systems used by hospitals and universities.
This earlier work demonstrates software-building breadth and longevity since 2007.
Medical training and clinical practice provide context for health-information workflows. Health-data-science training and research support evidence, evaluation and data-quality work.
Charity, foundation, venture and bilingual product experience support mission-led operations and English/Arabic delivery.
03 · Qualifications
Clinical, technical, analytical and delivery disciplines brought together.
04 · Published research
JAMIA · 2022 · 30(2):222–232 · FIRST AUTHOR
Musaab is first author of Translating and evaluating historic phenotyping algorithms using SNOMED CT, published in the Journal of the American Medical Informatics Association. The study translated and evaluated phenotyping algorithms across approximately 500,000 patient records and 276 disease phenotypes.
Read the published paper05 · Selected development evidence
Selected platforms and development demonstrators developed or technically led by Musaab show the practical AI, software-engineering and domain capabilities behind RefLens. They demonstrate the ability to turn complex requirements into working systems, while remaining examples rather than a limit on what RefLens can deliver.
Operational internal platform · employer delivery
An AI-assisted governance, risk and operations platform developed for a UK charitable trust and used in its internal operations. It brings together grant management, case management and wider organisational workflows.
Its public site describes 20 integrated modules and more than 75 AI-powered features. Musaab led its development from initial design through operational delivery, while a broader commercial edition continues to be developed.
View platform siteDevelopment-stage demonstrator · public development demo
A healthcare-workflow demonstrator covering clinical, appointment and operational patterns.
Demonstrates health-domain analysis, multi-tenant architecture, bilingual delivery and the ability to shape similar systems around a hospital's requirements.
View development demoDevelopment-stage demonstrator · public development demo
An AI-assisted clinical-coding demonstrator producing reviewable suggestions rather than autonomous coding decisions.
Demonstrates retrieval, terminology-aware workflows, auditability and responsible human review.
View development demoDevelopment-stage platform · public development demo
A research-reference verification platform using scholarly metadata, automated checks and human-review workflows.
Demonstrates applied AI, evidence-source integration, metadata comparison and end-to-end product engineering.
View development demoA client engagement may involve a bespoke solution, adaptation of a proven technical pattern, co-development of a reusable platform or a future hosted/SaaS route where that model is appropriate.
These examples demonstrate Musaab's product-building and delivery experience. They are not presented as completed RefLens client engagements or as finished RefLens SaaS offerings.
06 · Working approach
The goal is a useful, reviewable system that another team can operate.
Identify the decision, user and operational outcome that should improve.
Make assumptions, data quality, evaluation and limitations visible.
Document architecture and reasoning while the system takes shape.
Leave code, tests and documentation another team can use.
Work directly with the founder
Work directly with the person who would analyse the problem, shape the technical approach and lead the delivery.
Discuss a project