From AI models to community initiatives, I turn ambitious ideas into reality.
I'm a Data Science undergraduate at FAST NUCES (Class of 2027) based in Islamabad. My technical work spans NLP research, AI systems, and full-stack engineering — and outside of code, I'm equally at home behind a camera, at a desk, or advocating for injured wildlife.
As Founder of Parwarish, I'm building an AI-enabled mobile platform that puts user experience first, secured with national startup funding. I also co-founded a Wildlife Rescue initiative with the belief that compassion and technology aren't mutually exclusive.
I care deeply about building things that matter — whether that's a machine learning pipeline, a community fundraiser, or a hand-lettered piece of art.

Parwarish is a culturally grounded parenting support platform that helps Pakistani parents navigate modern parenting challenges through trusted guidance, community support, and expert-backed resources.




PakWild is a Pakistan-focused wildlife rescue, citizen science, and conservation platform. It is a digital ecosystem for wildlife rescue, biodiversity monitoring, and conservation collaboration across Pakistan.



Frontend-based society platform UI built for academic and community engagement with structured navigation and responsive layout.
Photography-centric frontend platform designed for visual storytelling, seamlessly connecting clients with professional photographers across multiple niches.
This platform is a full-stack movie recommendation and social network designed to transform movie discovery into an interactive, community-driven experience.
An R-based Multiple Linear Regression (MLR) analysis exploring the impact of behavioral, cognitive, and environmental factors (procrastination, focus span, and class attendance) on student CGPA at FAST University. Includes statistical modeling scripts, summary diagnostics, predictive tracking, and data visualizations.
Project is modeling and forecasting the national inflation rate of Pakistan (measured as the year-over-year percentage change in the Consumer Price Index, or CPI) using datasets spanning from 1980 to 2024, comparing a classical time-series model (ARIMA) against several machine learning frameworks (Lasso, Ridge, Elastic Net, SVR, and Random Forest).
NLP system for Urdu-English question answering using transformer-based XLM-RoBERTa architecture.
AI-based drone routing and delivery simulation using A* pathfinding, optimization, and demand forecasting models.
Scalable ANN-based retrieval system using HNSW indexing and distributed shard-based architecture.
Data-driven analysis addressing declining retail sales in the electronics department. Developed a complete pipeline featuring data cleaning, EDA, Linear Regression for spending trends, Decision Tree classification for customer types, and K-Means clustering for customer segmentation
A data-driven analysis of crime trends across the United States using FBI NIBRS data from 2000–2022. The project explores temporal, geographic, demographic, and offense-type patterns through Python-based data analysis and visualization techniques to uncover insights for law enforcement and policymakers.
The project focuses on recreating the classic "Plane Crash" game using the Unity engine. It aims to provide an improved experience over the traditional gameplay style by introducing smoother touch controls, optimized performance, and modern pixelated visuals.




From short-form social content to event coverage and cinematic reels, combining technical precision with a visual eye developed through years of photography and design work.
As Secretary Creatives at FCCS, led and supported initiatives that combined creative skills with meaningful social impact — organizing fundraisers, community outreach, and campus events that left a real mark.
Open to research, internships, freelance, and full-time roles. Also happy to connect over wildlife, art, or a shared love of Urdu NLP.
maimoonahanan.17@gmail.com