Welcome to my digital space

Ahmed Samir

AI Engineer • Researcher • Computer Vision Specialist

Undergraduate AI Researcher specializing in Computer Vision, 3D Spatial Reasoning, and Quantum Machine Learning — Kaggle Competitions Expert with a peak global rank of 768th (2nd nationally in Egypt), with multiple international hackathon victories and peer-reviewed research.

Explore My Work Get In Touch

Scroll to explore

About Me

I'm a fourth-year student at the Faculty of Computers and Artificial Intelligence, Cairo University (B.Sc. in Computers & AI, expected Jun 2027), specializing in AI with a focus on Computer Vision, 3D Spatial Reasoning, and Quantum Machine Learning.

I am a Kaggle Competitions Expert with a peak global ranking of 768th, and I am currently ranked 2nd nationally in Egypt. My Kaggle medals span computer vision, biology, and time-series forecasting — including the Image Matching Challenge (a 3D reconstruction pipeline built on DINOv2, ALIKED, and LightGlue), BirdCLEF+, CAFA 6 Protein Function Prediction, and CIMBTR (Top 10%).

My track record spans multiple international hackathon victories — including 1st places at the HERE Technologies and Alexandria Quantum Hackathons — and a Data Science internship at EFG Hermes engineering Reinforcement-Learning-driven high-frequency trading algorithms.

Hackathon Awards

AI & Research Projects

95.5%

SOTA YOLO mAP

768th

Peak Global Rank • Kaggle

Professional Experience

EFG Hermes

Data Scientist Intern • On-site, Egypt

Aug 2026 – Sep 2026

Engineered High-Frequency Trading Algorithms using Reinforcement Learning.
Designed the state-action space and reward functions for financial market environments.

SRSEG

Researcher • Scientific Research School of Egypt (Remote)

Dec 2024 – Apr 2026

Engineered and trained 3+ YOLO variants with novel backbones (Swin Transformer, CBAM) for Dentist X-Ray Object Detection, achieving a new state-of-the-art mAP of 95.5%.
Actively participate in weekly research meetings and webinars, presenting findings and shaping new research directions.

HERE Technologies

Data Scientist Intern • Hybrid

Sep 2025 – Dec 2025

Recruited into an internship following a 1st place win in the HERE Technologies Hackathon.
Worked on geospatial data with different database engines.
Worked with AWS services, Docker, and Linux.

National Telecommunication Institute

ML & NLP Trainee • Egypt

Feb 2025 – Jul 2025

Completed an intensive program in ML, DL, Agentic AI, and RAG on Huawei AI platforms.
Achieved a Kaggle Bronze Medal in the CIMBTR competition (Top 10%) as part of the graduation project.
Developed an educational tool using RAG and Agentic AI to assist math teachers in identifying student misconceptions.

Oberlin College

Quantum Mechanics Student Researcher • Remote

Mar 2023 – Jul 2023

Awarded Full Scholarship & a copy of the e-book "Where Research Begins".
Researched double-donor semiconductor Tellurium and wrote a 19-page paper.
Graded A, awarded 2 transferable credit hours, and evaluated as top 5% by Prof. Adhikari (Assistant Professor at Colgate University, PhD from Amherst College).

Education

🎓

B.Sc. in Computers & AI (Major in AI)

Faculty of Computers and Artificial Intelligence, Cairo University

Expected: Jun 2027

Coursework: Algorithms Analysis & Design, Supervised Learning, Autonomous Multiagent Systems, Formal & Natural Language Processing, Reasoning & Knowledge Representation

🏫

STEM High School for Boys – 6th of October

Physics Club President • Math Circle Vice President

Graduated: Jun 2023

Author of a Thermodynamics Course Book

Technical Projects

🔬

Image Matching Challenge 2025 — 3D Reconstruction & Point Cloud Generation Pipeline

Kaggle Bronze Medal • 2025

Engineered a scalable 3D reconstruction pipeline utilizing DINOv2 to extract global image descriptors, optimizing image pairing via similarity thresholding to bypass expensive exhaustive matching.
Implemented ALIKED to detect high-density local keypoints and integrated LightGlue for robust, high-accuracy feature matching across shortlisted image pairs.

🧩

Roboflow DETR (rf-detr) — Core Contribution

Open Source Feature Developer • 2026

Engineered strict configuration validation using Pydantic for the official Roboflow DETR repository, preventing silent model misconfigurations.
Implemented a custom configuration base class and unit tests to catch parameter typos, ensuring reliable fine-tuned weight loading for production users.

⚛️

QuYAML: A Token-Efficient Quantum Circuit Standard

Python Parser for AI-Driven Quantum Development • 2025

Designed a human-readable, token-efficient format (73% fewer tokens than Qiskit JSON) for quantum circuits, optimized for LLMs.
Developed the official Python parser (v0.2) converting .quyaml files into executable Qiskit QuantumCircuit objects.
Engineered a comprehensive test suite (100% coverage) using metamorphic testing to ensure mathematical correctness (unitary equivalence) for complex algorithms (QFT, VQE).
Benchmarked against OpenQASM and JSON, demonstrating 73% token reduction vs. JSON.

🚆

RbT: Return by Train — KAN with Self-Improving Loop

Top 15% Innovation & Creativity, 20th NU UGRF • 2025

Top 15% in innovation and creativity among 800+ projects from 10+ countries at the 20th NU UGRF.
Implemented a Kolmogorov-Arnold Network (KAN) for time series forecasting.
Designed a self-improving feedback loop (inspired by the Darwin Gödel Machine) allowing the model to retrain and adapt as new data arrives.

Awards & Competitions

🏅

Caspian Cup 2026 (Azerbaijan)

Faculty Nominee — Selected by FCAI to represent Cairo University internationally • 2026

🥇

HERE Technologies Hackathon

1st Place Winner • 2025

🥇

Alexandria Quantum Hackathon

1st Place Winner • 2025

🚀

NASA Space Apps

Global Nominee • 2025

🥈

Quantathon — AIMS Ghana

2nd Place Winner • 2025

🥈

Fawry AI Rising Gen

2nd Place Winner • 2025

🥉

Cairo AI Hackathon

3rd Place Winner • 2025

🏆

WIDS Hackathon

3rd Place Regionally • 2026

🥉

Kaggle — CAFA 6 Protein Function Prediction

Bronze Medal • 2026

🥉

Kaggle — Image Matching Challenge

Bronze Medal • 2025

🥉

Kaggle — BirdCLEF+