About Me

Hossam Elsherbiny

AI Engineer & Data Scientist

Professional Bio

I'm an AI Engineer and Data Scientist passionate about building intelligent systems that solve real-world problems. My expertise spans across Large Language Models, Generative AI, RAG Systems, Agentic AI workflows, Computer Vision, and NLP.

I specialize in turning complex AI research into production-ready applications. From designing multi-agent systems with LangGraph to deploying computer vision pipelines with FastAPI, I focus on building reliable, scalable AI solutions.

My work covers the full spectrum of AI engineering — from data collection and model training to API development and cloud deployment. I'm driven by the intersection of cutting-edge AI research and practical engineering.

Areas of Expertise

Programming Languages

PythonSQLRC/C++Java

Analytics Frameworks

NumPyPandasSciPyMatplotlibSeabornPlotlyStreamlitGradio

ML Frameworks

Scikit-learnTensorFlowKerasPyTorchXGBoostLightGBMCatBoost

Computer Vision (CV)

OpenCVYOLOViTMediapipeWav2Lip

Natural Language Processing (NLP)

TF-IDFTransformersWhisper

Generative AI

GPTBERTLlamaAutoEncoderGANsDiffusion ModelsHugging Face Transformers

Agentic AI

AI Agents (LangGraph, CrewAI, AutoGen)Advanced RAG (LangChain, Ragas)LangSmith

Time Series

ARIMAETSProphet

MLOps & Tools

GitDockerKubernetesMLflowJupyterTableauPowerBISpreadsheetsWeights & Biases

Big Data & Cloud

HadoopYarnHivePySparkKafkaAWS (EC2, S3)

Databases

SQLiteMySQLPostgreSQLMongoDBChromadbPgvector

Backend Frameworks

DjangoFastAPI

Work Experience

Oct 2024Present

AI Engineer

Freelance · Remote

  • Completed 100+ hours of freelance AI consulting and development, delivering high-impact solutions for international clients on platforms like Upwork.
  • Fine-tuned the FLUX.1-dev text-to-image model using LoRA adapters via ai-toolkit on cloud-hosted A100 GPUs, achieving high-fidelity custom character generation matching exact target concepts.
  • Built an end-to-end Egyptian Arabic speech recognition and structured named entity extraction pipeline utilizing a fine-tuned Whisper model, MARBERT tokenizers, and GPT-4o-mini.
  • Integrated MediaPipe skeleton landmark tracking and OpenCV for video-based gesture detection and audio-visual feature alignment.
PythonFlux.1WhisperLangChainGPT-4o-miniPyTorchOpenCVMediaPipeLoRADiffusers
Nov 2024Oct 2025

AI Engineer

Acrow · Cairo, Egypt

  • Extracted and processed data from Oracle databases to support time-series forecasting for over 3,000 SKUs.
  • Designed, trained, and evaluated forecasting models (SARIMA, XGBoost, CatBoost, Prophet, LightGBM, and neural networks), lowering forecast error from 80% to 55%.
  • Engineered an intuitive GUI-based executable application to automate the end-to-end data extraction, model training, and forecasting pipeline, decreasing the manual process time from 3 days into 1.5 hours.
  • Partnered with cross-functional teams to validate model outputs and integrate forecasts into the company’s planning processes, boosting profit by over 10.5%.
PythonOracleSARIMAXGBoostCatBoostProphetLightGBMNeural Networks

Education

Oct 2025Present

9-Month Professional Program

Information Technology Institute (ITI) · Mansoura, Egypt

Artificial Intelligence and Machine Learning Track

Oct 2020Jul 2025

B.Sc. in Computer Engineering

Faculty of Engineering, Zagazig University · Zagazig, Egypt

  • Graduated with Honors
  • Ranked 10th in Class

Interested in working together?