Senior AI/ML Engineer

I advise startups and enterprise clients on end-to-end agentic AI system architecture, technical strategy, and production deployment. I implement frontier LLM solutions for automation, incident reporting, research, data analysis, problem solving, and decision support.

I design reusable AI components, multi-agent retrieval workflows, and conversational agents with persistent memory, structured logging, and MCP-based tool integrations. I also implement RAG pipelines, fine-tuned models, multimodal AI workflows, and repeatable MLOps deployment practices; vector retrieval improvements increased candidate and document retrieval efficiency by 35%.

I architected and built a ChromaDB-based RAG retrieval system for knowledge-graph tasks in a government-funded France-Germany research project. The solution was containerized with Docker and deployed through Streamlit.

I developed multimodal NLP models for classification, summarization, generation, sentiment analysis, and named entity recognition, achieving over 94% classification accuracy and 95% entity-classification accuracy. I also provided technical leadership for computer-vision work using CNNs and OpenCV, achieving 98% accuracy in segmentation and structural classification.

During my PhD research, I designed feature-embedding and representation-learning models for multi-domain graph data using AI and NLP methods. I developed graph node and edge attribute models with more than 90% accuracy and proposed a multi-level GNN architecture that outperformed existing approaches by approximately 20%.

I built anomaly-detection and content-based recommendation models for predictive analytics and automated ML pipelines for training, evaluation, and optimization. I presented original research internationally, led internal knowledge-sharing sessions, and mentored junior researchers.

I led the development of a real-time data pipeline between a public web portal and PostgreSQL, supporting live data-stream updates and increasing operational throughput by 40% in a regulated public-sector environment.

I built a deep-learning model to predict COVID-19 spread patterns across countries and visualized findings using Plotly and Matplotlib. I collaborated on PostgreSQL infrastructure, ML pipeline automation, statistical reporting, data preprocessing, and analysis of complex user datasets.

I conducted research spanning web development, software development, data management, and data-pipeline preparation for analysis. I designed supervised and semi-supervised machine-learning models for pattern learning, prediction, and classification.

I contributed to a cross-team deep-learning model for crop-yield prediction, improving predicted crop-production outcomes by 35%.

Focused on artificial intelligence, graph representation, natural language processing, and deployment of machine-learning models for downstream tasks.

Graduated with Distinction (5.00/5.00). Focused on supervised and unsupervised machine-learning model design for IT optimization.

Graduated with First Class Honours (4.50/5.00).