Mohammad Othman

About Me

I'm the Founder & Lead AI Engineer at Transformer Labs, an AI consultancy delivering end-to-end LLM, RAG, voice AI, and automation solutions for clients across Europe, the Gulf, and the US. I design and deploy production RAG systems with hybrid retrieval and multilingual embeddings, build real-time voice AI agents with STT/TTS pipelines and telephony integration, deliver LLM fine-tuning engagements, and develop workflow automation and multi-agent systems. I work as a founding engineer, hands-on with clients on both long-term and short-term engagements.

Previously, I served as a Deep Learning Engineer at AUI™ (Augmented Intelligence), where I built LLM fine-tuning pipelines with DeepSpeed and Accelerate across multi-GPU nodes, deployed vLLM for high-throughput inference, introduced LiteLLM as a unified router across providers (Anthropic, Vertex AI, AWS Bedrock, Together), and productionized inference with Jenkins CI/CD including canary deployments and auto-rollback.

Before that, I worked as a Machine Learning Engineer at Jawwal, Palestine's largest telecom operator, operating at scale with 20M+ events/day. I built customer segmentation, CLV modeling, churn prediction, and credit scoring systems, along with drive-test analytics for network quality prediction. I also partnered with data engineering to stabilize a 200+ TB Oracle data warehouse and mentored a team of 4 engineers.

I also contributed as a Software Engineer at Foothill Technology Solutions, building applications with ASP.NET, C#, and REST APIs. My tenure as an NLP Scientist at the University of Aberdeen involved entity extraction, semantic similarity, medical NER, and low-resource language understanding research. I also gained experience as a Machine Learning Intern at Sky, supporting viewer-engagement modeling, ML pipeline monitoring, and A/B testing analysis.

I hold a Master's degree in Artificial Intelligence from the University of Aberdeen and a Bachelor's in Computer Engineering from the Eastern Mediterranean University.