Melat Mesele

Contributed to training and evaluation of large language models through data annotation, response ranking, and quality assessment. Reviewed outputs for factual accuracy and guideline alignment while following strict task constraints and evaluation rubrics.

Built a responsive, accessible UI for a language learning platform using Next.js and TypeScript. Implemented a reusable component library and integrated backend features like caching, optimistic updates, and error handling.

Developed secure, scalable payment gateways with Python and Flask to support reliable transaction processing. Built RESTful APIs integrating external payment services to enable multiple payment methods.

Built an end-to-end Java platform for large-scale time series analysis, anomaly detection, and interactive visualization. Implemented streaming ingestion with Spring Boot reactive streams plus MQTT/Kafka, and supported versioned feature storage with Redis and PostgreSQL.

Worked with a team of developers to strengthen end-to-end delivery practices and consistent deadline execution. Used Jira, GitHub, and Slack to improve communication and project tracking.

Led a team of 6 backend engineers to build AfroChat’s backend and improve delivery workflows. Migrated from Flask + MongoDB to FastAPI + SQLAlchemy (async), added CI/CD with GitHub Actions and Docker, and created Python tools that reduced container build times by 40%.

Developed an interactive coding practice and exam platform enabling practice, online exams, and instant feedback. Built the frontend with Next.js and the backend with NestJS and MySQL, including an NLP-based plagiarism detection system to compare submissions.

Earned a Bachelor of Science in Software Engineering at Addis Ababa University. Coursework included web development, computer architecture, OOP, machine learning, natural language processing, and mobile development.