Skip to main content
Hossam El-Kharbotly

About Me

I'm an AI Engineer & Automation Specialist at Dar, one of the world's top-ranked multinational engineering consultancies, where I build production-grade AI systems at the intersection of enterprise knowledge management and cost intelligence. My work is grounded in intensive research — I benchmark models, frameworks, and architectures rigorously before committing to production, because the wrong choice at scale is expensive to reverse. My expertise spans agentic workflows with multi-step reasoning, RAG architectures (GraphRAG, Knowledge Graphs, Vector DBs), on-premise LLM serving, and LLM optimization — with a track record of delivering measurable impact in production environments.

I specialize in building intelligent systems that autonomously reason, use tools, and self-correct. My technical foundation combines on-premise LLM deployment (vLLM, Qwen, Gemma, fine-tuning), multi-agent orchestration (LangGraph, 10+ node architectures), knowledge graph engineering (Neo4j, vector search), computer vision (medical imaging with 93% recall, explainable AI), NLP, and production ML deployment (FastAPI, Docker). My journey includes securing first place in two university research competitions and progressing from R&D roles to deploying production AI systems at the scale of global engineering organizations.

When I'm not coding or studying, you'll find me playing tennis 🎾, biking 🚲, or enjoying a game of football ⚽.

Education

B.Sc. in Electronics and Computer Engineering

Nile University

2019 - 2024

  • • Cumulative Grade: Very Good
  • • Presented 6 posters in the Undergraduate Research Forum (UGRF), winning first place in two.
  • • Full Year Research Graduation Project Grade: Excellent - A

Experience

Dar

Full-Time

AI Engineer & Automation Specialist

Feb 2026 – Present

  • • Built enterprise-wide knowledge base with on-premise LLMs (Qwen, Gemma via vLLM), Knowledge Graphs, and multi-source RAG pipelines
  • • Deployed 10+ node LangGraph multi-agent system for cost engineering — natural language queries over SQL databases, BoQ analysis, and project cost estimation
  • • Researching Recursive Language Models (RLM) and knowledge graph architectures for hierarchical document parsing and accumulative institutional memory
  • • Designed multi-source ingestion pipelines (SharePoint, databases, archives) deployed on-premise and Azure

Edulga

Full-Time

AI Engineer

August 2025 – January 2026

  • • Built production AI platform with GraphRAG, Neo4j knowledge graphs, Pinecone vector search, and LangChain orchestration
  • • Achieved 85% performance improvement via parallel processing, HTTP/2 pooling, and intelligent caching
  • • Reduced infrastructure costs by 50% with FastAPI/Docker microservices while maintaining sub-second query latency
  • • Implemented 97+ production tests, LLM monitoring, and cost tracking for API optimization

Al-Farid Scan

Part-Time

R&D Machine Learning Engineer

March 2024 – March 2025

• Improved MRI classification algorithms by 20%, enhancing diagnostic accuracy through innovative machine learning techniques.

The Central Administration for Water Resources and Irrigation

Internship

Data Analyst

Dec 2023 – Jan 2024

  • • Database management (querying, searching, etc.).
  • • Assessing the performance of telemetry systems and developing solutions for improvement.

Technical Expertise

AI & Machine Learning

  • Large Language Models (LLMs) - Fine-tuning, Prompt Engineering, LLM Optimization
  • On-Premise LLM Serving - vLLM, Qwen, Gemma, model benchmarking and evaluation
  • Vision Language Models (VLMs) - Document parsing, engineering drawing analysis
  • RAG Systems - Retrieval Augmented Generation, GraphRAG Architectures
  • AI Systems and Agentic AI - LangChain, LangGraph, LlamaIndex, Multi-Agent Workflows
  • Natural Language Processing (NLP) - BERT, Transformers, Sentence Embeddings, Arabic NLP
  • Computer Vision - Medical Imaging, Grad-CAM, Explainable AI (XAI)
  • Vector Databases - FAISS, Pinecone, Weaviate, Qdrant
  • Knowledge Graphs - Neo4j, Graph Neural Networks, Accumulative Knowledge Design
  • Recursive Language Models (RLM) - Hierarchical semantic parsing, tree-structured document representation, semantic composition for nested technical content

Frameworks & Tools

  • ML Frameworks - PyTorch, TensorFlow, Scikit-learn, Hugging Face
  • LLM Orchestration - LangChain, LangGraph, LlamaIndex, Sentence Transformers
  • LLM Serving - vLLM, FastAPI
  • MLOps and Deployment - LangFuse, Docker, FastAPI
  • Cloud Platforms - AWS, Azure, Google Cloud
  • Databases - SQL, MongoDB, Vector DBs (Weaviate, Qdrant, FAISS, Pinecone)
  • Version Control and CI/CD - Git, GitHub Actions, GitLab CI
  • Data Processing - Pandas, NumPy, OpenCV

Completed Courses

Machine Learning Specialization

Coursera - DeepLearning.AI

Deep Learning Specialization

Coursera - DeepLearning.AI

PyTorch: Deep Learning and Artificial Intelligence

DeepLearningCourses.com

OpenCV Course – Full Tutorial with Python

FreeCodeCamp

AI Programming with Python Nanodegree Program

Udacity

Achievements

Band Score 7 - C1

IELTS Academic Exam

Campus Director - 37 teams

Hult Prize at NU

Finalist - Shortlisted Participant

European Union EBTIKAR Competition

First Place - Two times

Undergraduate Research Forum (UGRF)

Finalist - Shortlisted Participant

UNDP AfriClimate Inclusive Innovation Bootcamp

Presented 6 Research Posters

Undergraduate Research Forum (UGRF)