Transforming data into insights, complexity into clarity and Building ML pipelines
Computer Science student passionate about machine learning, natural language processing, and building intelligent systems that solve real-world problems. Based in Frankfurt, Germany.
Parameter-efficient fine-tuning (LoRA) of OpenAI's Whisper model targeting regional dialects of Amharic, a low-resource language spoken by 80 million people. Features custom script normalization pipelines and low-VRAM deployment setups to accurately parse crowdsourced African regional audio datasets.
GitHub Repo
Building a comprehensive home lab infrastructure with private cloud capabilities for self-hosting services, running ML experiments, and learning enterprise-level DevOps and Cyber security practices.
View Architecture
Contributing to AI safety seminar by investigating alignment problems, robustness challenges, and ethical considerations in deploying large language models and AI systems.
Read DraftExtracting meaningful patterns from complex datasets using statistical analysis and visualization.
Building intelligent systems that understand and process human language with state-of-the-art models.
Automated data collection and extraction from web sources for analysis and research.
Creating robust server-side applications and RESTful APIs with modern technologies.
Sharing knowledge and helping others master mathematics and programming concepts.
Advanced sentiment analysis using DimABSA (Dimensional Aspect-Based Sentiment Analysis) with transformer architectures for nuanced emotion detection and aspect-level sentiment extraction. Accepted & Published at SemEval 2026.
Interactive dashboard analyzing Tesla's financial performance using real-time SEC data. Features comprehensive stock price tracking and revenue analysis with dynamic visualizations.
View Project
Complete NLP pipeline featuring web scraping, UIMA framework integration, WhisperX audio transcription, and RESTful API for seamless data processing and analysis.
View ProjectReal-time news article classification system using NLP and machine learning. Built with Streamlit for instant predictions and user-friendly interface.
View ProjectCurrently based in Frankfurt, Germany. Open to opportunities in data science, machine learning, and software development.