Mohammad Aqueel

I am an AI/ML engineer with over two years of experience designing and deploying end-to-end AI and LLM systems for production use. I focus on building scalable, real-time applications including voice agents, semantic search, and histopathology analysis platforms.

At Softsensor AI I developed deep learning models for cell and nuclei detection on very large WSI/DICOM images, created internal DICOM handling libraries, and integrated ML models via FastAPI into a web annotation platform with real-time editing.

I built hybrid semantic search and domain-aligned QA systems using vector databases (Qdrant) and re-ranking techniques (ColBERT), plus multilingual handling and intent-based classification to improve retrieval relevance and user experience.

I engineered voice-based conversational AI solutions using streaming and speech-to-speech tooling, plus microservices, Docker orchestration, WebSocket and queue systems for long-running inference, and managed data with Postgres, MongoDB and vector stores to enable analytics and human-in-the-loop workflows.