Azka
AI Developer & Researcher
MSc Computer Engineering @ Gachon University | BS IT (1st Class Honors) @ Air University
AI developer building production ready LLM systems with a focus on security, privacy, and reliability.
Research Focus: LLM & RL to enhance security | Fintech & Sustainability.
Engineering Philosophy: Turn ideas into working systems that people can trust.
"I work at the intersection of research and deployment—turning LLM ideas into secure, reliable, real-world systems."
Based in Seoul, South Korea | Origin: Islamabad, Pakistan
Research Areas: Agentic AI, RL Based agents, and ESG & Sustainability
About Me
I'm an AI Developer with 1st Class Honors in BS IT from Air University, Islamabad, currently pursuing a Master's degree in Computer Engineering at Gachon University, South Korea.
My research focuses on LLM Security, FinTech Innovation, and Sustainable AI Systems, with a passion for developing socially responsible and ethically aligned AI solutions.
I've led cross-functional teams across 3+ international research collaborations and have contributed to cutting-edge research in AI security, sustainability, and financial technology.
Professional Experience
Research Experience
- • Conducting research on LLM Security with focus on adversarial robustness and safe deployment
- • Investigating privacy-aware guardrails for generative models under resource-constrained settings
- • Exploring cloud-scalable AI architectures with emphasis on security and privacy
- • Contributing to cutting-edge research in AI security, privacy, and ethical AI deployment
Resulted Papers:
Projects:
- • Contributing to AI-driven sustainability frameworks in financial systems, aligned with ESG and SDG goals
- • Authored research paper on PTSD detection using NLP techniques, achieving 97.56% accuracy
- • Focusing on risk mitigation in AI models for credit scoring, fraud detection, and ethical automation
- • ML-based PTSD prediction through textual data analysis in healthcare systems
Resulted Papers:
Projects:
- • Writing comprehensive review paper on Sustainable AI Approaches for Edge Devices
- • Conducting practical research on energy-efficient AI techniques and low-power algorithms
- • Collaborating with international teams to evaluate environmental sustainability in AI
- • Implementing sustainable AI approaches to reduce energy consumption in edge computing
Resulted Papers:
Projects:
Industry Experience
- • Developed RAG-based AI assistant for real-time business data communication and analysis
- • Built automated reputation management system with advanced sentiment analysis capabilities
- • Annotated 20k+ reviews to train aspect-based review analyzer with high accuracy
- • Optimized Generative AI models (Phi-3, GPT-4.0) reducing operational costs by 60%
- • Created speech-to-speech reservation system using NLP and speech recognition
- • Developed AI-powered software for automated business marketing and customer engagement
- • Built text-to-text Generative AI applications using GPT-3.5 for multiple domains
- • Created Django-based RESTful API endpoints for 9 different AI-powered modules
- • Developed multi-specialty medical chatbot with API integration for healthcare applications
Key Achievements:
Links:
- • Developed robust iOS applications (Dialer & Messenger) using Swift and Objective-C
- • Created ML-based message classifier using LSTM, increasing accuracy by 70%
- • Led AdTech initiative to predict user call behavior for targeted advertising optimization
- • Researched CDR data usage for predicting user call patterns and marketing insights
Key Achievements:
Links:
- • Designed and delivered comprehensive ML course outlines for international students
- • Trained 50+ students worldwide with 95% course satisfaction rate
- • Enhanced students' knowledge through engaging practical content and hands-on projects
- • Created interactive learning materials and assessment frameworks
Key Achievements:
Links:
Internships
- • Built ML models for rare event classification in audio data, achieving 90% accuracy
- • Created and annotated a custom audio dataset for forensic analysis
- • Developed a web-based application for audio forensics and analysis
Links:
- • Developed and deployed iOS apps with integrated ML features for enhanced user experience
- • Spearheaded app design with high user engagement focus
- • Implemented machine learning capabilities within mobile applications
Links:
- • Worked on startup innovation strategies with focus on experimentation and risk analysis
- • Contributed to entrepreneurship projects through strategic planning and analytics
Links:
- • Developed a hyper-casual Unity game 'Rob it All' in a multidisciplinary team
- • Trained in 3D modeling, Unity, Blender, and gaming pipeline fundamentals
Links:
Fellowships & Training
- • Participated in technical workshops and lectures from leading ML researchers across MENA
- • Topics included Responsible AI, LLMs, and Federated Learning
Links:
- • Selected for a competitive program bridging academia and industry through mentorship
- • Worked on a machine learning-based final year project under supervision of senior engineers
Links:
- • Completed intensive training sessions in career development and applied AI
- • Collaborated on a real-world ML project under industry mentorship
Links:
Education & Certifications
Focus Areas: Generative AI, ML Systems, Cloud Architecture
Ongoing Research: LLM Security, FinTech Innovation, Sustainable AI Systems
Relevant Coursework: AI, Data Science, Software Engineering, ML
Final Year Project: PTSD Prediction using NLP (97.56% accuracy)
Certifications
Featured Projects
Context-aware healthcare assistant using large language models. FastAPI-powered AI chatbot delivering expert responses across cardiology, psychiatry, and dentistry using specialty-specific prompt templates.
A Stable Diffusion-based simulation of cosmetic effects on facial images. Web-based tool simulating aesthetic treatments (e.g., Botox, fillers) on uploaded images. Users can visualize cosmetic outcomes in real time.
Achieved 97.56% accuracy using NLP techniques on clinical text data. Built a text-based neural model to detect PTSD from patient records, contributing to mental health diagnosis and early intervention.
A sentiment analysis dashboard for restaurant reviews using text classification. System for review classification, sentiment analysis, and dashboard visualization with automated processing capabilities.
Retrieval-augmented chatbot for restaurant owners to interact with business data, generate reports, and analyze sales. Built for US market with comprehensive business intelligence features.
An AI-powered tool that generates, refactors, and optimizes code snippets for multiple languages. Developer productivity API offering endpoints for code generation, bug fixing, and explanation.
Technical Skills
Research Collaborations & Team Work
Throughout my research journey, I've had the privilege of collaborating with talented researchers and developers across multiple international projects. These collaborations have shaped my understanding of global AI challenges and solutions.
Bio-AI Lab Research Team, Tromsø, Norway
Collaborative research on sustainable AI approaches for edge devices with international teams at University of Tromsø.
FinTech Research Collaboration, University of East London
AI-driven sustainability frameworks in financial systems with Dr. Fahimeh Jafari and research team.
AI Development Team, AIO Startup
Cross-functional AI development team working on RAG-based systems and generative AI applications.
MENA ML Winter School 2025, Qatar
International machine learning conference with researchers from across MENA region at QCRI.
"Great achievements are born from great collaborations. These experiences have enriched my perspective on global AI research and development."
What Colleagues & Supervisors Say
Dr. Fahimeh Jafari
Senior Lecturer, University of East London
University of East London
"Azka has demonstrated exceptional research capabilities in our FinTech AI collaboration. Her work on PTSD detection using NLP techniques achieved remarkable 97.56% accuracy, showcasing her deep understanding of both technical implementation and real-world applications."
Prof. Research Supervisor
Research Director, Bio-AI Lab
University of Tromsø, Norway
"Working with Azka on sustainable AI approaches has been incredibly productive. Her ability to bridge theoretical research with practical implementation, especially in edge computing environments, has contributed significantly to our publications."
Senior Developer
Lead AI Engineer
AIO Startup
"Azka's contribution to our AI development team has been outstanding. She successfully optimized our generative AI models, reducing operational costs by 60% while maintaining high performance. Her technical expertise in RAG systems is exceptional."
Team Lead
iOS Development Manager
FITFLEX
"During her time with us, Azka developed robust iOS applications and created ML-based message classifiers that improved accuracy by 70%. Her leadership in AdTech initiatives and technical innovation made a significant impact on our projects."
Let's Connect
Ready to collaborate on innovative AI research or discuss opportunities? I'd love to hear from you!
More About Me: