Kirollos Ehab Portfolio

Who Am I?

I'm Kirollos Ehab, Machine Learning & AI Engineer specializing in LLM-driven systems, and intelligent QA automation. Experienced in building agentic workflows, and ML-based solutions in industrial environments at Siemens EDA. Strong background in deep learning, data-driven system design, and scalable software development, with experience in developing end-to-end ML pipelines and production-ready tools.

Siemens EDA, AI/ML QA Engineer July 2024 - Present

Enhanced model performance through hyperparameter optimization (Optuna, GridSearch), improving prediction accuracy and inference efficiency
Designed a test prioritization framework using ML to rank regression test cases, reducing execution cost while maintaining high failure detection coverage
Developed agentic QA workflows integrating LLMs, tools, and execution pipelines to automate test case generation and analysis at scale
Integrated Model Context Protocol (MCP) servers to enable tool-augmented LLM reasoning, improving context-awareness and automation reliability
Refactored legacy Perl/Python QA automation pipelines, improving maintainability, execution performance, and debugging visibility
Built internal chatbot tools using prompt engineering to assist QA/RD teams with technical queries and workflow acceleration
Authored technical documentation for ML pipelines and QA systems, improving onboarding and reproducibility

ASU Racing Team, Deep Learning Researcher November 2023 - July 2024

Collaborating with a team of talented individuals to design and implement innovative machine learning models for autonomous vehicles
Applying deep learning algorithms to improve the team’s performance in various autonomous vehicle competitions

Cellula Technologies, ML Engineer Intern November 2023 - December 2023

Led a high-performing team in the development of Taxi Fare Prediction, achieving significant enhancements in accuracy and efficiency
Used Classification techniques as DecisionTrees, KNN, RandomForest & Logistic Regression
Used Regression techniques as DecisionTreesRegressor, Linear Regression, RandomForestRegressor & XGBoostRegressor
Achieved high accuracy & low MSE while working on the tasks

Edges for Training, Embedded Systems Diploma July 2023 - November 2023

Worked under supervision of Engineer Mohamed Tarek
Mastered the basic concepts of Embedded Systems, demonstrating a solid understanding of core principles
Mastered the fundamental concepts of Embedded C
Interfaced with AVR Micro-controllers, developing comprehensive drivers for various peripherals
Acquired knowledge of Real Time Operating Systems (RTOS)

Telecom Egypt, AI Trainee July 2023 - August 2023

Learned About Neural Networs & Worked with Keras, Tensorflow & NumPy
Knew CNN Architecture & difference between ANN & CNN
Used Pre-Trained models like MaskRCNN for Object detection & instance segmentation