Looking for a job

omar mohamed ahmed

@omarmohamedahmed

Message

AI Engineer at Deakin University (A2I2) building production RAG systems that improve retrieval precision by 20%.

Australia

Message

At Deakin University (A2I2), I fine-tune open-source LLMs and deploy them through Docker and AWS SageMaker. This work cut inference latency by 15% and improved task accuracy by 10% across classification and summarization benchmarks.

I built a production RAG system with LangChain and LlamaIndex, combining hybrid retrieval with cross-encoder re-ranking. It serves more than 500 students and researchers and improved retrieval precision by 20% over a dense-only baseline.

Through Upwork (Independent Freelance), I scoped, built, and shipped production AI systems for legal, healthcare, and e-commerce clients. My projects included document Q&A, semantic search, multilingual text classification, and speech recognition pipelines.

My Ph.D. research at Deakin University (A2I2) focused on LLM alignment, privacy, memorization, and multilingual NLP. I've published work at NAACL, EMNLP, EACL, and LREC.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

Deakin University (A2I2)

Feb 2025 - Present (1 year 8 months)

Fine-tuned open-source LLMs using PEFT/LoRA, cutting inference latency by 15% and improving task accuracy by 10%. Built a production RAG system with hybrid retrieval and re-ranking, improving retrieval precision by 20%.

LLM Fine tuning PEFT LoRAs RAG Lang Chain LlamaIndex Docker AWS Sagemaker CI CD Ragas TruLens

AI Engineer

Upwork

May 2022 - Jun 2023 (1 year 1 month)

Independently scoped, built, and shipped 10+ production AI systems for clients in legal, healthcare, and e-commerce. Built end-to-end RAG applications and deployed fine-tuned models as REST APIs on AWS.

RAG Faiss Pinecone fastAPI AWS Whisper Wav2Vec Python

Data Scientist

iNetworks

Jan 2020 - Apr 2022 (2 years 3 months)

Built a real-time NLP-driven recommendation system using sentence embeddings and text classification, contributing to a 25% increase in user engagement. Engineered automated web scraping and preprocessing pipelines, cutting manual data preparation effort by 40%.

NLP Recommendation Systems Embeddings Text Classification Web Scraping Data Pipelines

Undergraduate Research Assistant

RSSCI Lab, Helwan University

Oct 2019 - Dec 2021 (2 years 2 months)

Designed and ran model evaluation experiments across computer vision and NLP tasks using PyTorch and TensorFlow. Led large-scale data acquisition and preprocessing for image and text datasets.

PyTorch TensorFlow Computer Vision NLP Data Engineering

Education

Degrees, certifications, and relevant coursework

Deakin University

Doctor of Philosophy, Artificial Intelligence

2023 -

Pursuing a PhD in Artificial Intelligence with research on LLM alignment, privacy, memorization, and multilingual NLP. Published at NAACL, EMNLP, EACL, and LREC.