Projects are structured as mini case studies highlighting the problem context, technical approach, and
outcome.
ApplAI - AI Assisted Job Application Automation Platform
March 2026
Built an AI-driven workflow to automate job application tasks such as CV tailoring, document generation, and structured extraction from job descriptions.
What I built
- Implemented LLM-based pipelines to transform unstructured text into role-specific outputs.
- Designed modular workflow stages to support scalable and repeatable automation.
- Applied prompt engineering strategies to improve consistency and quality of generated artifacts.
Generative AI
Prompt Engineering
Workflow Automation
ARPA Meal Planner - AI Assisted Meal Planner
February 2026
Developed an intelligent meal planning application that optimizes meal selection using nutrition, cost, and user preference constraints.
What I built
- Designed data models and transformation pipelines for food dataset analysis.
- Implemented filtering and recommendation-oriented planning workflows.
- Structured the architecture for future AI-driven recommendation extensions.
Data Pipelines
Optimization Logic
Scalable Design
Crime Insights in Dallas - ML Classification, Clustering and Outlier Detection
Algonquin College · November 2025
Analyzed a 1.4M+ record police incident dataset using CRISP-DM to extract insights from noisy and imbalanced data.
What I built
- Built pipelines for classification, clustering, and anomaly detection.
- Performed encoding, feature engineering, dimensionality reduction, and model tuning.
- Evaluated Decision Tree models via grid search and cross-validation while diagnosing imbalance effects.
Machine Learning
CRISP-DM
Data Mining
HubSpot to Zoho Custom Integration
Call Center Studio · November 2023
Engineered middleware in Google Apps Script to synchronize deal and line-item data from HubSpot to Zoho CRM across departments.
What I built
- Implemented API calls with `UrlFetchApp` and payload transformations with XML and JSON parsing.
- Mapped and pushed processed records into account, potential, and sales-order objects.
- Improved cross-system data reliability for operational reporting and workflow continuity.
Google Apps Script
REST APIs
CRM Integration
Telecom Customer Churn Prediction and Decision Support System
Bogazici University · January 2022
Built an end-to-end churn analytics workflow from data preparation to model optimization on a real-world telecom dataset.
What I built
- Handled class imbalance with SMOTE, ADASYN, and class-weighted learning strategies.
- Trained and tuned CatBoost, XGBoost, and LightGBM with stratified validation and Optuna.
- Improved ROC-AUC performance from 0.74 to 0.84.
Python
Scikit-learn
Optuna
Developing an Android Application for Crowdsourcing: Spotify
Bogazici University · January 2023
Created a Kotlin Android app that leverages Spotify audio analysis and an ExpressJS endpoint to deliver chord data to users.
What I built
- Designed Android flows to collect Spotify Track IDs and display processed output.
- Built an ExpressJS integration layer to connect the app with existing open-source processing logic.
- Integrated Firebase for authentication and backend data support.
Kotlin
Android
ExpressJS
Firebase
Email Signature Generator
December 2021
Built an internal signature generator page to help new team members produce consistent, branded email signatures.
What I built
- Designed and implemented reusable HTML/CSS signature structure and controls.
- Added lightweight JavaScript form behavior for generating and copying signature output.
HTML
CSS
JavaScript