Looking for a job

abdalla abdelsamie

@abdallaabdelsamie

Applied AI Engineer at WallTech Soft building LLM and RAG features for the CargoWare logistics platform.

At WallTech Soft, I build and support AI and LLM features for the CargoWare logistics SaaS platform. I design LLM and RAG solutions for logistics and customer workflows, and integrate AI services with backend systems using Python and FastAPI.

At TP Cargo, I fine-tuned LLaMA with RLHF (DPO) on 50k+ samples and created Flask APIs over 100k+ documents. The work improved retrieval by 15% and reduced latency by 20%.

At Desay SV, I worked on malware detection using PyTorch and Jetson Orin, and CAN/Ethernet anomaly detection. The systems achieved 92% malware detection accuracy and improved anomaly detection performance by 18%.

My work also includes computer vision and real-time control services in manufacturing, alongside embedded software for vehicle ECUs. I hold a Master's in Mechanical Engineering from Northwestern Polytechnical University, grounding my software work in a software–hardware foundation.

Experience

Work history, roles, and key accomplishments

Current

Applied AI Engineer / Technical Solution Consultant

Current

WallTech Soft

Sep 2025 - Present (1 year 1 month)

Build and support AI/LLM features for WallTech logistics SaaS (CargoWare). Design and implement LLM/RAG solutions for logistics and customer workflows, set ML/LLM standards, and integrate AI services with backend systems.

Python LLM RAG Lang Chain Hugging Face fastAPI Logistics SaaS

Backend Developer (AI Vision & Control)

Intelligent Manufacturing / Automation

Jun 2025 - Sep 2025 (3 months)

Built FastAPI services for real-time motion control and OpenCV defect detection, with MySQL/MES traceability. Achieved ~15% lower motion latency and improved bonding accuracy and visual QC.

Python fastAPI OpenCV MySQL MES

Senior AI / ML Engineer (Business Intelligence)

TP Cargo

Oct 2024 - May 2025 (7 months)

Provided LLM + RAG support, fine-tuned LLaMA with RLHF (DPO) on 50k+ samples, and built Flask APIs over 100k+ docs, achieving +15% retrieval and −20% latency. Developed LSTM models for freight pricing on 80k+ rates with real-time quote APIs, improving prediction accuracy by 10%.

PyTorch Hugging Face LLaMA RLHF DPO RAG Flask Time Series ML

R&D Embedded Software / AI Developer

Desay SV

Aug 2022 - Aug 2024 (2 years)

Developed malware detection using PyTorch MLP on 200k+ samples on GPU/Jetson Orin, and CAN/Ethernet anomaly detection with XGBoost. Applied OpenCV on vehicle camera logs, and worked on TEE secure boot and SOME/IP for ECUs, achieving 92% malware detection accuracy and +18% anomaly detection performance.