Mahinour Mohammad

Applied AI Engineer specializing in production-grade LLM systems, agentic RAG architectures, and security automation. Strong in business logic thinking and end-to-end solution architecture. Experienced in hybrid retrieval (Dense + Sparse + ColBERT + RRF), asynchronous pipelines, real-time anomaly/fraud detection, and RAG evaluation.

Delivered scalable, low-latency AI solutions for document intelligence and smart-meter analytics, processing millions of data points with 99%+ integrity and measurable reliability improvements.

AI and Machine Learning Engineer at Advanced Engineering Consultations (ADVEC) (2024-07 – Present)

Architected and deployed production-grade RAG systems, smart meter data collection, and IoT automation solutions with focus on data integrity and reliability improvements.

• Architected and deployed a production-grade RAG system with custom ingestion pipelines, hybrid search (Dense + Sparse + ColBERT), reranking (RRF), and Qdrant.
• Engineered a multi-threaded DLMS/COSEM data collection system for integrating 500+ smart meters, with OBIS code mapping for supported profiles.
• Built data processing pipelines with real-time failure/fraud detection, processing millions of readings (99%+ data integrity).
• Developed 50+ RESTful Flask APIs for metering, configuration, and remote load control, automating billing and cutting manual effort by 90%.
• Integrated Hugging Face LLM for automated financial Excel categorization, reducing report errors by 75%.
• Built secure FastAPI + MQTT monitoring for 20+ sensors with Pydantic validation and dashboard integration.
• Reduced API latency by 40% via asynchronous FastAPI pipelines (background tasks + ThreadPoolExecutor), Redis caching, and efficient embedding management.
• Built RAG evaluation pipelines (RAGAS + custom logging) to monitor grounding, model drift, and hallucination rates, improving response quality in production.
• Contributed to product planning and authored PRDs, evolving Lingudeep from a translator to a full document-intelligence platform (8-person cross-functional team).
• Developed object detection model for a smart irrigation system, fine-tuning an SSD model to identify plant diseases from Raspberry Pi camera images.

AI & Machine Learning Diploma in Artificial Intelligence & Machine Learning – Information Technology Institute (ITI) (2023-10 – 2024-06)

Bachelor of Science in Communication & Electronics Engineering – Port Said University (2018-10 – 2023-06)