# ~/about/bio.py
class NayabGull:
def __init__(self):
self.role = "AI/ML & App Developer"
self.education = "BS Information Technology @ Rawalpindi Women University (CGPA 3.72/4.0)"
self.focus = ["Machine Learning", "AI-Integrated Apps", "Flutter & Android"]
self.currently_building = "DreamScape — AI-integrated Android app"
self.open_to = ["AI/ML internships", "Mobile dev opportunities", "Research collabs"]
def philosophy(self):
return "Ship real, deployed products — not just prototypes."I'm a final-year BS IT student who builds end-to-end — from EDA and model training to shipped, live products. I've deployed two full-stack AI web tools (FraudScope, Casefile), built an LLM comparison dashboard, run multiple ML pipelines on Kaggle, and shipped Firebase-backed Flutter/Kotlin mobile apps. I'm first author on an in-progress IEEE-formatted research paper on cross-domain log anomaly detection, and I won RWU IdeaFest 2026 for a tech-for-good project.
Languages
AI / Machine Learning / Data Science
Mobile & Frontend
Backend & Data Storage
Tools & Deployment
> FraudScope — Privacy-First Fraud Detection Tool
A live, full-stack fraud/anomaly detection tool for financial ledgers. All detection math (outlier detection, duplicate/round-number/threshold-skirting checks, timing anomalies, vendor concentration) runs locally on the server with pandas/numpy — an LLM only ever receives anonymized, aggregated statistics, never raw transaction data. Includes a consent-gated AI chatbot for report Q&A, AI narrative summaries, PDF report export, and a one-click Offline Mode.
stack: [Python, Flask, pandas, numpy, Groq LLM API, reportlab, JS/HTML/CSS]
status: live
demo: fraud-scope.onrender.com
repo: github.com/nayab-gull-it/FraudScope> Casefile — AI-Powered Resume Intelligence Platform
A live, full-stack AI web app that scores a resume against any job description, flags missing keywords, and rewrites the resume to better match the role — using only the user's real experience. Includes PDF/DOCX export and AI-generated cover letters.
stack: [Python, Flask, SQLite, Groq LLM API, HTML/CSS/JS]
status: live
demo: casefile-resume-agent.onrender.com
repo: github.com/nayab-gull-it/casefile-resume-agent> LLM Comparison Dashboard
Interactive dashboard comparing 70+ LLMs (GPT, Llama, Deepseek, Claude, Gemini, Mistral +) from 8 providers across speed, latency, context window, benchmark scores (MMLU, Chatbot Arena), pricing, and energy efficiency.
stack: [Python, Streamlit, Pandas, Plotly]
status: live
demo: llm-comparison-dashboard.streamlit.app
repo: github.com/nayab-gull-it/LLM-Comparison-dashboard> Student Grade & GPA Prediction
End-to-end ML pipeline on student academic data — EDA, preprocessing, classification and regression modeling, plus SHAP explainability to interpret individual predictions.
stack: [Python, Pandas, Scikit-learn, SHAP, Matplotlib, Seaborn]
results: "92.5% accuracy (classification) | R2 = 0.943 (regression)"
repo: github.com/nayab-gull-it/student-grade-gpa-prediction> Traffic Flow Prediction
Regression modeling to predict traffic flow from road type, day type, weather, and time, with hyperparameter tuning via GridSearchCV.
stack: [Python, Scikit-learn, GridSearchCV]
results: "R2 = 0.986 (Gradient Boosting Regressor)"
repo: github.com/nayab-gull-it/traffic-flow-prediction> House Price Prediction
EDA and regression modeling comparing Linear, Ridge, Lasso, Random Forest, and tuned Gradient Boosting.
stack: [Python, Scikit-learn, GridSearchCV]
results: "R2 = 0.913 (tuned Gradient Boosting)"
repo: github.com/nayab-gull-it/house-price-prediction-ml> Netflix Content Analysis (EDA)
Exploratory data analysis on Netflix's content catalog — content type trends, release patterns, and country-wise distribution.
stack: [Python, Pandas, Matplotlib, Seaborn]
repo: github.com/nayab-gull-it/netflix-content-eda> DreamScape — AI-Integrated Android App
An Android app that uses AI to interpret dream journal entries — built end-to-end, not yet published to the Play Store.
stack: [Kotlin, Firebase, Gemini API, SQLite]
architecture: offline-first
status: built, unpublished
repo: private> Smart Community Safety App
A Flutter-based mobile app (academic semester project) promoting community safety through incident reporting, an interactive unsafe-zone map, community alerts, an SOS emergency mode, and an analytics dashboard.
stack: [Flutter, Firebase Auth, Firestore, Provider, Google Maps]
features: [incident reporting, safety map, SOS mode, analytics]
repo: github.com/nayab-gull-it/smart-community-safety-app> Pet Shop Management System
A desktop-based system for managing pet inventory, customer records, billing, and sales transactions.
stack: [C#, MySQL]
features: [inventory management, billing, sales tracking]- role: Flutter Developer Intern
company: Tech4Edges
period: July 2026 – Present
- role: Android Developer Intern
company: Sqrock IT Solutions
period: August 2026 – Present
- role: Flutter App Internee
company: Developers Corporation
period: Dec 2025 – Feb 2026
focus: Flutter apps with Firebase auth & real-time database integration
- role: Support and IT Internee
company: PNY Training Institute
period: March 2025 – April 2025
focus: LMS management, customer follow-ups, day-to-day IT support
- role: IT Support
company: 4th International Conference of Sciences
focus: On-site technical support during conference sessions"Cross-Domain Log Anomaly Detection Using One-Class SVM: A Unified Model Across HDFS, BGL, and Thunderbird Systems"
First Author · Under Preparation, 2026
A unified One-Class SVM approach to detecting anomalies across three distinct system log datasets, formatted for IEEE submission.
| 🏆 | RWU IdeaFest 2026 — Winner — tech-for-good project, built with a teammate |
| 🎓 | CGPA 3.76/4.0 |
Certifications
Open to AI/ML internships, mobile dev opportunities, and research collaborations.
Still learning, still building — one commit at a time.