AI/ML & Computer Vision Engineer
Aleena Shakil
sees what models miss.
7th-semester Software Engineering student (CGPA 3.71/4.00) who built a production-style banknote recognition and counterfeit-detection pipeline at the State Bank of Pakistan — training classifiers in PyTorch, extracting text with PaddleOCR and OpenCV, and shipping deployment-ready tooling with FastAPI.
Trained on real problems,
not just benchmarks.
I work at the intersection of applied computer vision and deployable software — the space between a model that scores well in a notebook and a pipeline that holds up against real, messy, scanned input.
At the State Bank of Pakistan, that means classifying banknote denominations across thousands of images, correcting perspective and stripping background noise before OCR ever sees a note, and documenting every iteration so the model's reasoning survives the handoff — not just its weights.
Outside of that, I teach programming labs, mentor through Google Developer Groups on Campus, and build small end-to-end projects to keep the fundamentals — data preprocessing, evaluation, deployment — sharp.
Where the training happened.
AUG 2026
Summer Intern
State Bank of Pakistan · IT & Project Management Dept- Built and evaluated a PyTorch classification model to recognize denominations across 7,500+ banknote images, using accuracy and confusion-matrix analysis to guide iterative improvement.
- Developed an OCR text-extraction pipeline with PaddleOCR and OpenCV, applying perspective correction and background removal to improve read accuracy on scanned notes.
- Contributed CV preprocessing steps to a counterfeit-detection pipeline, supporting testing against real and simulated counterfeit samples.
- Collaborated with a cross-functional team via Git, documenting model iterations for project handoff.
JAN 2026
Teaching Assistant
Sir Syed University of Engineering & Technology- Supported students in programming and ICT lab sessions, debugging code and clarifying core concepts in real time.
- Reviewed and graded assignments, providing targeted feedback that helped students fix recurring logic errors.
PRESENT
Member
Google Developer Groups on Campus (GDGOC)- Participated in technical workshops and AI-focused learning sessions.
- Collaborated with peers on software development and applied AI projects.
Pipelines, end to end.
Bangladeshi Banknote Recognition & Counterfeit Detection
End-to-end computer vision pipeline combining a PyTorch classifier with PaddleOCR text extraction — handling perspective correction, background noise, and multi-denomination classification across a 7,500+ image dataset.
Layered a counterfeit-detection workflow on top of the classifier using targeted CV preprocessing.
SmartFood AI
AI-powered meal recommendation system with a custom preprocessing pipeline for nutritional data.
Integrated the Gemini API to generate personalized food recommendations from user input.
Pet Shelter Management System
Database-backed application for managing pet records, adoptions, and medical history.
Selected for the SSUET Project Exhibition.
Attendance Management System
Attendance-tracking application with reporting dashboards for monitoring attendance trends over time.
The stack, by layer.
- Python
- Java
- SQL
- PyTorch
- TensorFlow
- Scikit-learn
- Model eval & feature engineering
- OpenCV
- PaddleOCR
- Image preprocessing
- Object detection fundamentals
- Pandas
- NumPy
- Matplotlib
- Seaborn
- FastAPI
- React
- Git
- Jupyter / Colab
- MongoDB
- MySQL
Foundations.
B.S. Software Engineering
Certifications
- Machine Learning Specialization — Andrew Ng, Coursera
- SQL for Data Science — Simplilearn
- Introduction to Modern Database Systems — Saylor Academy
Achievements
- Merit Scholarship recipient at SSUET
- Summer Intern at the State Bank of Pakistan (SBP)
- Selected for SSUET Project Exhibition
- Active member, Google Developer Groups on Campus (GDGOC)
Have a dataset worth training on?
Open to internships, research collaborations, and applied CV/ML roles. The fastest way to reach me is email.