terminalStart

loading profile … ok

Clinical data · Bioinformatics · Research infrastructure

Reliable research data, from the bench to federated scale.

I’m Mogahid Sabeel, a Junior Clinical Data Manager with Alleanza Contro il Cancro. I bring a background in bioinformatics, HPC and workflow automation to one question: how do we make research data accurate, traceable and ready for responsible use?

3yrs
biomedical research data at IRCCS San Raffaele
2007
first role in scientific data analysis
4
working languages: Arabic, English, Italian, French
Fig. 01 — Data becoming dependable. Irregular inputs are checked, harmonised and traced until they are ready for research use.

readme.txtNow

Clinical data for a federated research network.

Since July 2026 I have worked as a Junior Clinical Data Manager with Alleanza Contro il Cancro, operationally based at IEO – European Institute of Oncology, contributing to Health Big Data: a national programme connecting the Italian IRCCS research networks.

  • Clinical-data curation, structuring, standardisation and validation
  • Systematic quality control: missing data, ranges and inconsistencies
  • Data dictionaries, semantic harmonisation and interoperability
  • Coordination across clinical, research, data-management and platform teams

Views on this site are my own and do not represent any employer or programme.

Clinical-data quality is not simply a final check. It is part of the research infrastructure.

— from my two-month reflection, September 2026

PortfolioWork

Evidence, labelled honestly.

Selected work from previous and ongoing practice. Every card says what stage the work actually reached — delivered, strategy, or practice — and what is deliberately left out.

DeliveredSan Raffaele · 2023–26

Structured capture & approval workflows

Power Platform applications that turned quotation, approval and sample-tracking steps into structured, traceable records for research facilities.

  • Power Apps
  • Power Automate
  • SharePoint
DeliveredSan Raffaele · 2023–26

Reproducible NGS workflows & HPC support

Bioinformatics and NGS workflows in Python, Bash and Snakemake, plus hands-on support for researchers running and troubleshooting jobs on HPC.

  • Snakemake
  • Python · Bash
  • HPC · Conda
Strategy · proposalSan Raffaele

AI infrastructure strategy for biomedical research

Translated genomics, imaging and AI-research workloads into a phased, governable architecture: pilot, scale, federate.

  • HPC / AI architecture
  • Roadmap
  • Governance
Delivered · Phase 1 pilotSan Raffaele

Digital workplace sovereignty pilot

Assessed EU-centric and open-source alternatives, then deployed and documented a working multi-service pilot on an EU-hosted server.

  • Docker
  • Matrix · Jitsi
  • Nextcloud · Authentik
Working practicePersonal · ongoing

Source-traceable knowledge operations

A local-first knowledge system that separates source facts, proposals and decisions — with closeout records and hand-over rules for AI-assisted work.

  • Obsidian
  • Markdown
  • AI agents
Learning labSynthetic data only

Clinical-data & infrastructure practice labs

Synthetic-only practice: SQL quality checks, a FHIR sandbox for Patient-resource workflows, and an OpenStack + Ansible lab for account automation.

  • SQL
  • HAPI FHIR
  • Ansible

case.txt

Method wizardMethod

Make every step checkable.

Step 1 of 4

  1. Map the real workflow

    Start from inputs, people, decisions and hand-offs — before choosing a tool or changing a process.

  2. Make checks explicit

    Turn assumptions into visible rules: completeness, ranges, consistency — and who resolves each exception.

  3. Keep provenance

    Record where each value came from, which version of a rule applied, what changed and why.

  4. Design for the next person

    Leave runbooks, data dictionaries and decision logs that let someone else repeat the work without me.

Proposals are not decisions

Draft, proposed, sent and approved are different states — and I label them that way.

Unknown stays unknown

Missing evidence is recorded as missing, never quietly filled in or averaged away.

Synthetic first

Prototypes and tests run on synthetic data before anything touches real records.

Sources over memory

Every important claim points back to a source someone else can check.

Copying…Journey

From animal genetics to clinical data — one thread.

Different fields, same question each time: can we trust the data enough to act on it?

  1. 2026 →

    Junior Clinical Data Manager

    Alleanza Contro il Cancro · operationally based at IEO, Milan

    Clinical-data quality, harmonisation and interoperability within the Health Big Data programme.

  2. 2023 – 2026

    Researcher — bioinformatics, HPC & digital research support

    IRCCS Ospedale San Raffaele · CSTOR & COSR

    Power Platform applications, NGS workflows, HPC support and digital research infrastructure, working across scientific, operational and technical teams.

  3. 2022 – 2023

    Assistant instructor — programming & computational thinking

    CodEng · Varese

    Teaching programming fundamentals — and learning to explain technical ideas simply.

  4. 2017 – 2020

    MSc, Medical, Veterinary & Pharmaceutical Biotechnologies

    Università degli Studi di Padova

    Bioinformatics, omics, molecular pathology and epidemiology.

  5. 2007 – 2015

    Research scientist — animal genetics & data analysis

    Agricultural research · Sudan

    Research data collection, analysis and reporting — the first lessons in data quality.

  6. 2001 – 2006

    BSc (Hons), Animal Science

    University of Gezira · Sudan

MixerToolkit

Skills, with honest levels.

Clinical & research data

  • Data quality & validation rulesWorking proficiencyCurrent role focus
  • Data dictionaries & harmonisationWorking proficiencyCurrent role focus
  • GDPR-aware data handlingWorking proficiencyGDPR training completed; privacy-by-design habits
  • HL7 FHIRStudying / learningStudied; synthetic sandbox
  • OMOP CDM · OHDSIStudying / learningStudied and compared
  • CDISC · EDC · SQLStudying / learningLearning with synthetic studies

Bioinformatics & compute

  • NGS workflowsDeep practiceDemultiplexing, QC, alignment, variant calling
  • Snakemake · NextflowWorking proficiencyReproducible pipelines
  • Python · BashWorking proficiencyWorkflows, parsing, automation
  • HPC · job scheduling · CondaWorking proficiencyResearcher support and troubleshooting
  • RStudying / learningBasic

Automation & structured capture

  • Power AppsDeep practiceForms, validation, LIMS-style apps
  • Power AutomateDeep practiceApproval and notification flows
  • SharePoint · ExcelDeep practiceData structures, permissions, sites
  • VBA · JavaScriptStudying / learningBasic

Infrastructure & strategy

  • Linux · virtualisationWorking proficiencyEveryday administration
  • Docker multi-service stacksWorking proficiencyPhase 1 collaboration pilot
  • Technical strategy & platform evaluationWorking proficiencyRoadmaps, option analysis, decision briefs
  • Source-traceable documentationDeep practiceRunbooks, decision logs, closeouts
  • OpenStack · AnsibleStudying / learningLab environment

Languages

  • Arabicnative
  • EnglishC2
  • ItalianC2
  • FrenchB1

MessageContact

Let’s talk about research data.

For professional exchanges on data quality, reproducible workflows and research infrastructure, LinkedIn is the best place to reach me.