ali alavi
@alialavi
I build and evaluate machine learning systems for LLM uncertainty, reinforcement learning, and AI agents.
ZimbabweWhat I'm looking for
I'm the sole engineer on Zodu LLC's TRACE project, designing and running PyTorch pipelines for model training, data attribution, and uncertainty evaluation in large language models. I also built a Gymnasium environment for Dungeon Crawl Stone Soup, ran PPO training with Stable-Baselines3 and CUDA, and fixed silent observation-pipeline defects that prevented learning.
As a Computer Science master's student at the University of Messina, I continue work in machine learning and Python development. My background includes CNN-based herbarium image classification, database teaching assistance, and responsive web development with on-page SEO.
Experience
Work history, roles, and key accomplishments
TRACE: Training Attribution for Confidence Estimation
Only engineer on TRACE, a research project on training data attribution for uncertainty estimation in large language models. Work in progress toward publication.
I design and run the full experimental pipeline in PyTorch: model training, attribution runs, and evaluation. (Methodology withheld pending publication.)
Teaching assistant for the Database Design Principles course. Supported students on their course projects and graded assignments, papers, and exams.
Built responsive websites for client projects, working with the development team through the full build cycle. Worked on on-page SEO, improving content structure, meta tags, and page performance to help pages rank higher in search.
Education
Degrees, certifications, and relevant coursework
Università degli Studi di Messina
Master's degree, Computer Science
Master’s student in Computer Science at the University of Messina, continuing my work in machine learning and Python development
Birjand University of Technology
Bachelor's degree, Computer Engineering
2018 - 2023
Completed my Bachelor's in Computer Engineering with coursework in programming, algorithms, machine learning fundamentals, and digital systems.
My final thesis project focused on applying Convolutional Neural Networks (CNNs) for Herbarium image classification, which introduced me to deep learning concepts and practical model development.
Availability
Location
ZimbabweAuthorized to work in
Job categories
Interested in hiring ali?
You can contact ali and 90k+ other talented remote workers on Himalayas.
Message aliGet matched with your dream remote job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
