AI Research Engineer with less than a year in Data Science & LLM Fine-tuning
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Data Science graduate from PUCIT with hands-on experience in machine learning, deep learning, and LLM fine-tuning. Developed AI systems ranging from computer vision to LLM security. Seeking an entry-level AI/ML Engineer role.
Punjab University College of Information Technology (PUCIT)
Bachelor of Science · Data Science
August 1, 2022 – June 30, 2026
Punjab University of Computing and Information Technology
Teaching Assistant (Information Security)
January 1, 2026 – May 31, 2026
Lahore, Punjab, Pakistan
LLM Shield
June 1, 2025 – June 1, 2026
Developed LLMShield, a 4-module AI security platform detecting prompt injection, data poisoning, embedding vulnerabilities, and C code exploits in LLM-based applications, improved attack detection accuracy across attack categories. Fine-tuned BERT for data poisoning detection on safe and poisoned dataset, additionally fine-tuned Qwen 2.5 and LLAMA 3.2 7b for research and comparative analysis. Conducted vector embedding security testing on databases, identifying retrieval manipulation risks, documented mitigation strategies aligned with OWASP Top 10 for LLMs (2025).
Face Mood Reader with Quote & Music Recommender
June 1, 2025 – June 1, 2026
Built a real-time facial emotion recognition system using CNN and OpenCV from webcam input. Developed a Flask REST API and React frontend to provide mood-based quote and music recommendations. Implemented data preprocessing and model-inference pipeline for seamless real time interaction.
Brain Tumor Classification using MRI & Deep Learning
June 1, 2025 – June 1, 2026
Developed a deep learning model to classify brain tumors from MRI scans using TensorFlow/Keras on the Figshare dataset (2,870 images), achieving 94% classification accuracy. Implemented data preprocessing and augmentation techniques to improve model generalization across 4 tumor types.
Data Analytics Hackathon (PUCON'25)
January 1, 2025 – December 31, 2025
Analyzed traffic stop datasets to surface statistically significant patterns of demographic disparity in policing, presented findings to a panel of industry judges and placed in top 4 out of 12 teams.
Cultural Fit Analysis
The candidate's academic projects, particularly 'LLM Shield' and 'Brain Tumor Classification', demonstrate a strong interest in cutting-edge AI research and practical application. The 'Face Mood Reader' project shows initiative in building end-to-end systems. The hackathon participation indicates a collaborative and competitive spirit. The teaching assistant role suggests a willingness to contribute to a learning environment. The overall profile aligns well with a research-oriented, innovative culture.
Soft Skills & Operational Fit
The candidate's project descriptions indicate an ability to work on complex, multi-module systems and present findings (hackathon). The teaching assistant role suggests communication and mentorship skills. However, without direct assessment data, further evaluation of soft skills and operational fit is limited.