Adonay D. Gebremedhin
Software | ML | Computer Vision
Computer Engineering
I'm a computer engineering student focused on AI research, especially computer vision, multimodal learning, and agentic systems. I like turning research ideas into working prototypes, with recent work spanning medical imaging MLLMs, event-cameras, and visual object tracking and segmentation.
Work Experience
Ajman University
Feb. 2025-Present
Technology Transfer / Software Intern, Innovation Center
Support university-incubated startups by leading MVP development and providing technical support, helping early-stage teams turn ideas into functional products.
Jun. 2025-May 2026
Vice President, Computer Engineering Club
Coordinated technical events, workshops, and competitions while managing committee members and increasing engineering student engagement in practical portfolio building.
Jun. 2025-May 2026
Cabinet Member, IEEE Student Chapter
Managed website development, CI/CD, and digital presence while supporting consistent content coordination and platform growth.
Eastern Al Maha, EMCO
Jun. 2025-Aug. 2025
Computer Vision Intern
Trained and deployed three computer vision models for automated retail-shelf auditing, managed inference and deployment environments, and improved model accuracy through dataset curation and augmentation.
Smallcap.ai
Jun. 2024-Oct. 2024
Artificial Intelligence / Backend Intern
Led integration across AI, backend, frontend, and database workstreams; built Flask, React, MongoDB, and MySQL applications; and deployed PyTorch models for resume analysis and visual web scraping.
Education
Ajman University
BSc. Computer Engineering
CGPA: 4.0
UAE Ministry of Education fully-funded scholarship
Research
Under Review
Promptable Continuous Event Stream Segmentation
The 37th British Machine Vision Conference (BMVC) 2026
CheXGround: Anatomical Region Tokens for Grounded Longitudinal Chest X-ray Interpretation
Zayed University SRC 2025
Enhancing Human-AI Interaction through Multi-Agent Explainability
Multi-agent LLMs
Projects
PlanetWatch
A geospatial intelligence system for automated, scalable disaster response using vision and machine learning workflows.
Battle of Nets
An experimental deep learning project comparing modern models against established architectures.
CrewTailor
A web platform for building no-code, multi-agent LLM systems from scratch, created for the Gemini API Developer Competition.
End-to-End Multi-Object Localization and Path Planning
A robotics pipeline combining object localization and path planning for autonomous agents in a grid-based environment.
Awards
Third Place, Khalifa University Smart Mobile App Contest
PlanetWatch: geospatial intelligence for automated, scalable disaster response.
First Place, IEEE UAE Software Engineering Project
Radiance: a computer vision application for visually impaired users with object recognition, scene recognition, and agentic support.
Honors Assembly, Ajman University
Recognized for outstanding academic performance.