IFS Emerging Technologies · PwC Middle East

Building AI products that turn emerging technologies into real-world impact.

I'm Fatima Alabdullah, a BSc Artificial Intelligence student at King's College London with experience across AI prototypes, data dashboards, frontend development, and business analysis.

AI @ King's College London Business Analyst @ Saudi AZM PyTorch · RAG · LLMs · Dashboards First Class Honours

View Projects ↓ Open CV ↗ Contact Me

Scroll to explore

About

Technically capable.
Business-driven.

I am a BSc Artificial Intelligence student at King's College London with First Class academic standing and experience across AI development, business analysis, frontend development, and digital transformation.

My work focuses on turning emerging technologies into usable products — including clinical decision-support tools, computer vision systems, financial dashboards, and web platforms. I understand both the technical architecture and the business problem it needs to solve.

I am based between London and Saudi Arabia, and I am deeply invested in the digital transformation happening across the Middle East region — which is exactly where I want to grow.

View Projects LinkedIn ↗

1st

First Class Honours

King's College London

0.2%

Qimam acceptance rate

50 selected from 20,000+ applicants

50%

Reconciliation time reduced

Health Endowment Fund operations

130+

Alumni supported

through responsive web platform

Featured Projects

From prototype
to real product.

Each project represents a full journey: identifying a problem, building the technical solution, designing a usable interface, and thinking through adoption and impact.

AI Clinical Decision-Support · Prototype

Lung Cancer Risk Diagnosis Assistant

An AI-powered clinical decision-support prototype combining symptom analysis, chest X-ray interpretation, and RAG-based explanation generation to support lung cancer risk assessment.

My Role Co-led UI design and frontend architecture. Designed a conversational Flask-based interface, implemented user flows for symptom input and result presentation, integrated voice input via OpenAI Whisper, and partially contributed to the CNN inference pipeline, RAG retrieval flow, and LLM-based explanation outputs.

Python PyTorch DenseNet121 ChromaDB RAG Mistral AI Flask Whisper SentenceTransformers

Demonstrates ability to build explainable, user-facing AI systems combining computer vision, RAG, and LLMs — with a focus on real clinical usability.

GitHub ↗ Live Demo ↗

Demo note: Space may take ~5 min to wake. Login: doctor1 / password123

Plant Disease Detection interface preview

Computer Vision · ML Deployment

Plant Disease Detection System

A computer vision system that detects plant diseases from leaf images using a CNN-based image classification pipeline, deployed as an interactive Gradio interface on Hugging Face Spaces.

My Role Supported data preparation, model development, evaluation, and deployment. Built the interactive Gradio interface for image upload and prediction display. Used ClearML for experiment tracking.

Python PyTorch CNNs torchvision ClearML Gradio Hugging Face

Shows ability to move from model development to a polished, usable AI product interface — the full ML engineering arc.

GitHub ↗ Live Demo ↗

Demo note: This Hugging Face Space can take up to 5 minutes to load.

Confidential

Business Intelligence · Shefa.sa Client Delivery

Health Endowment Fund Financial Dashboard

While working at Saudi AZM, I supported the client Health Endowment Fund through the Shefa.sa platform project. This in-house Metabase dashboard enabled real-time tracking of SAR 50M+ in donations, case completion, and fund allocation in Saudi Arabia.

My Role Translated business requirements into technical workflows, defined dashboard specifications, coordinated with engineering teams across Saudi Arabia and Pakistan, and supported QA and onboarding.

Metabase SQL Jira Business Analysis Dashboard Spec

Enabled real-time tracking of SAR 50M+ in donations. Reduced reconciliation time ~50%. Improved cross-team communication across Saudi Arabia and Pakistan.

Confidential Code

Confidentiality note: Dashboard details and source code are restricted due to client requirements for the Health Endowment Fund.

Frontend Development · Web Platform

Rashid Alrashid Alumni Website

A responsive alumni website designed to improve digital accessibility, engagement, and communication across a 130+ member alumni community.

My Role Led a team of three. Designed and developed responsive UI components using React and Tailwind CSS. Coordinated with stakeholders and managed team delivery.

React Tailwind CSS JavaScript HTML/CSS

Improved digital accessibility and engagement for a 130+ member alumni network. Demonstrated product ownership and cross-functional team coordination.

Confidential Code Live Site ↗ Full Report ↗

Website note: The website is currently under maintenance. Please check the full report for complete details.

Research

Applied ML reports with
measured outcomes.

Two in-depth reports showing my approach to data analysis, model selection, and evidence-based conclusions under real ML constraints.

Regression Modeling · Applied ML

Outcome Prediction on a Simulated Diamond Dataset

Built and compared regression models on a 10,000-row training set with 30 features, then showed the target was simulated and weakly tied to classic diamond pricing variables.

Depth emerged as the dominant signal (corr ~ -0.411), while price and carat had near-zero relationship with outcome.
Best model was tuned XGBoost with R² = 0.4736, outperforming linear regression and random forest.
Feature importance indicated depth contributed ~57% of split decisions, with remaining signal limited by high irreducible noise.

XGBoost Feature Engineering RandomizedSearchCV R² 0.4736

Read Report ↗ GitHub ↗

Active Learning · Applied ML

TypiClust on a Budget: Baseline vs Modified

Reproduced TypiClust on CIFAR-10 and proposed a stronger variant to improve low-budget sample selection under cold-start active learning.

Proposed improvements: L2-normalized embeddings, KDE-based typicality, and overclustering (K = 500) with diverse cluster selection.
Linear Probe results improved clearly across budgets: 49.97% at budget 10 and 72.38% average vs 64.09% baseline TypiClust.
In fully supervised training with tiny label budgets, performance remained noisy, highlighting low-data instability.

SimCLR TypiClust KDE Typicality CIFAR-10

Read Report ↗ GitHub ↗

Experience

Where technical
meets business.

From delivery and analysis to engineering execution, I focus on measurable outcomes, stakeholder alignment, and production-ready solutions.

Aug 2024 – Jul 2025

Business Analyst

Shefa.sa / Saudi AZM · Riyadh, Saudi Arabia

Partnered with product and engineering teams to turn complex operational requirements into delivery-ready workflows, dashboards, and automation outputs.

Business Analysis Data Dashboard Client Delivery Cross-team QA

View outcomes, scope, and tools

Led delivery of a confidential Metabase dashboard for the Health Endowment Fund, tracking SAR 50M+ in donations in real time.
Defined requirements and translated stakeholder needs into technical tasks for distributed teams across Saudi Arabia and Pakistan.
Reduced reconciliation time by ~50% through improved operational workflows and structured reporting.
Supported QA, issue triage, and onboarding to improve adoption and handover quality.

Aug 2023 – Jan 2024

Frontend Developer

Abdulmonem Alrashed Humanitarian Foundation · Saudi Arabia

Led frontend execution for an alumni platform, balancing UX quality, accessibility, and stakeholder expectations under delivery deadlines.

React Tailwind CSS Responsive UI Team Leadership

View outcomes, scope, and tools

Designed and built responsive UI components to improve cross-device usability and accessibility.
Improved engagement and communication experience for a 130+ member alumni network.
Coordinated delivery across three contributors and aligned output with non-technical stakeholders.

Jan 2024 – Present

Chief Technology Officer

Smout · Saudi Arabia

Leading Smout's technology strategy as a youth-led initiative advancing science and space innovation in Saudi Arabia through digital platforms and national education programs.

Leadership Tech Strategy Space Education Program Design

View outcomes, scope, and impact

Led digital delivery and technical planning for Smout initiatives, including collaboration workflows, platform operations, and program execution.
Represented Smout in the Saudi Leadership Society (SLS) Challenge by Misk, contributing to innovation-focused national leadership tracks: Misk SLS report.
Served as a project judge in the Saudi National Space Program in collaboration with Starlight, evaluating youth projects and technical presentation quality.
Designed and ran structured online learning tracks for high-school students using Zoom and Slack, covering space fundamentals, project development, and presentation readiness.

Skills

Full-stack AI builder.

From model training to user-facing interface to business delivery — I operate across the full technical and product stack.

AI / Machine Learning

PyTorch CNNs scikit-learn YOLO TorchVision TorchXRayVision SentenceTransformers RAG LLM APIs Mistral AI OpenAI Whisper

Data / Analytics

Pandas NumPy SQL Metabase ChromaDB SQLite ClearML

Web / Product

React Tailwind CSS Flask Gradio Streamlit HTML/CSS JavaScript

Tools / Workflow

Git / GitHub Docker Hugging Face Spaces Jira REST APIs

Languages

Arabic Native

English Fluent

Why PwC

Emerging technologies
at scale.

⚙️

Why Emerging Technologies

My interest in emerging technologies comes from seeing how AI and digital tools move from technical prototypes into real products that solve business and social problems. Across my projects, I have worked on the full journey: identifying a problem, building the technical solution, designing a user-facing interface, and thinking carefully about adoption, usability, and measurable impact.

🌍

Why PwC Middle East

PwC Middle East's work in digital transformation, AI, data, and large-scale technology adoption across the region strongly aligns with the kind of work I want to grow in — combining technical capability with consulting, business understanding, and measurable client impact. The Middle East context, particularly Vision 2030 and the accelerating digitalisation of both public and private sectors, is where I believe emerging technologies can have the most transformative effect.

"I don't just build AI systems — I think about who uses them, why they matter, and how they create value beyond the proof of concept."

Get in Touch

Let's build technology
that creates measurable impact.

Open to conversations about the PwC Middle East IFS Emerging Technologies Internship and how my experience can contribute to your clients and teams.

Email Me ↗ LinkedIn ↗ GitHub ↗ Open CV ↗

fatimaxy.alabdullah@gmail.com · London / Saudi Arabia · LinkedIn · GitHub