Ahmed Moghazy — AI Engineer

PwC ETIC — AI Engineer

June 2026 — Present

• Worked in a team building an internal plugin that makes PDF and Microsoft Office documents compliant with PwC standards.

Advanced Computer Technology (ACT) — Software Developer (AI & Backend)

Oct 2025 — June 2026

• Built a multi-agent hotel concierge in LangGraph for guest requests and policy questions, with 95% intent classification accuracy.
• Built hierarchical RAG pipeline on PGVector for hotel policy documents; recursive retrieval cut hallucinations on long queries by 30%.
• Connected the agents to Oracle OPERA PMS via REST APIs to create and track guest service requests, cutting manual front-desk entry by an estimated 40%.
• Built voice agents on LiveKit and ElevenLabs: a multilingual hotel agent connected to OHIP / OPERA Cloud, and an Egyptian Arabic ordering agent connected to Oracle Simphony POS.
• Containerized the entire multi-agent stack using Docker, streamlining the CI/CD pipeline and reducing deployment downtime across internal testing environments.

Digital Egypt Pioneers Initiative (DEPI) — Machine Learning Engineer Trainee

Oct 2024 — Apr 2025

• Completed the Microsoft Machine Learning Engineer track, focusing on designing scalable and production-ready machine learning solutions using Microsoft technologies.
• Built strong foundations in statistics, linear algebra, Python, data preprocessing, visualization, classical ML, deep learning, NLP with attention mechanisms, and computer vision using transfer learning.
• Applied Azure AI Fundamentals and Azure AI Engineer Associate concepts, with hands-on exposure to MLOps tools including MLflow and Hugging Face.

Orange Egypt — Data Engineer Intern

Aug 2024 — Sep 2024

• Automated and optimized ETL workflows for voucher data using Apache NiFi and dbt, improving data quality and reliability.
• Explored pipeline orchestration and scheduling with Apache Airflow to increase automation and reduce manual data-handling effort.

ValU — Data Science Intern

July 2024 — July 2024

• Analyzed company growth metrics and customer behavior to generate data-driven strategic insights.
• Designed and delivered interactive Power BI dashboards visualizing KPIs, improving executive decision-making.
• Built and evaluated ML models (Random Forest, XGBoost) for customer lifetime value prediction, achieving an AUC of 0.70.
• Contributed to a customer-facing chatbot by unifying vendors and geographic regions using Named Entity Recognition (NER) for accurate offer retrieval.
• Optimized LLM prompts for intent extraction and analyzed OpenAI API usage, reducing inference costs by approximately 30% while maintaining response quality.

Advanced Computer Technology (ACT) — Data Science Intern

Aug 2023 — Sep 2023

• Conducted EDA on large datasets and presented insights to stakeholders.
• Automated web scraping pipelines using BeautifulSoup.

AI-Powered Tech News Aggregator (n8n, LLMs)

• Built an automated news aggregation pipeline using n8n to ingest RSS feeds from multiple tech sources on a scheduled basis.
• Implemented an LLM-based classification and ranking system to categorize articles, generate why-it-matters insights, and score news importance.
• Designed daily and weekly workflows to store structured results in Google Sheets and deliver curated HTML email summaries of top-ranked articles.

AutoML Pipeline (scikit-learn)

• Engineered a modular AutoML pipeline in Python (scikit-learn) with 4 stages: preprocessing, feature selection, model selection, and hyperparameter optimization.
• Synthesized AutoML foundations (model selection, HPO, evaluation) into the system architecture and experiment plan.
• Implemented and benchmarked Grid Search, Random Search, and Bayesian Optimization for HPO on tabular ML tasks, analyzing trade-offs in performance and compute cost.

RAG Chatbot (LangChain + FAISS + Ollama)

• Built a retrieval-augmented chatbot with LangChain + Ollama Gemma-3 (12B) and a Gradio UI for real-time Q&A.
• Crawled up to 30 LangChain doc pages (BeautifulSoup); chunked via RecursiveCharacterTextSplitter.
• Embedded with all-MiniLM-L6-v2 and indexed in FAISS; boosted relevance using MultiQueryRetriever + ContextualCompressionRetriever; added SerpAPI + Python REPL tools.

ExpenSum — Smart Expense Tracker (React + Spring Boot + JWT + LLM)

• Developed a full-stack expense tracker: React frontend and Spring Boot backend secured with JWT authentication.
• Integrated Mistral (via Ollama) to convert natural-language inputs into structured expense entries, enabling single-step data capture.

Star-Schema Data Warehouse (SQL, ETL)

• Designed a star schema (fact/dimension) for a recommendation dataset.
• Automated daily CSV loads via scheduled SQL jobs (ETL).

Programming

Python (Advanced), SQL, C++, JavaScript, HTML/CSS

Machine Learning & AI

Scikit-learn, TensorFlow/Keras, Classical Machine Learning, Deep Learning, NLP, Computer Vision, Feature Engineering, Model Evaluation, Hyperparameter Optimization

LLMs, Agents & Voice

LangChain, LangGraph, Multi-Agent Systems, Retrieval-Augmented Generation (RAG), MCP, Prompt Engineering, Claude API, Claude Code, OpenAI API, Gemini, Hugging Face, Ollama, LiveKit Agents, ElevenLabs, FAISS, Vector Databases

Data Analysis & Visualization

Pandas, NumPy, Matplotlib, Seaborn, Power BI, Exploratory Data Analysis (EDA)

Data Engineering & MLOps

Apache Airflow, Apache NiFi, dbt, ETL Pipelines, MLflow, HDFS, Workflow Orchestration

Databases

PostgreSQL, PGVector, MySQL, Star Schema Design, Data Warehousing

Cloud & Tools

AWS (Bedrock), UiPath, Azure AI Fundamentals, Huawei Cloud, Git/GitHub, Docker, Linux, Azure Machine Learning Concepts

Core Computer Science

Data Structures & Algorithms, Operating Systems, Distributed Systems, Database Management Systems

Soft Skills

Technical Communication, Problem Solving, Leadership, Analytical Thinking, Adaptability