ayesha ayesha

I've built and deployed production AI systems spanning LLM tool calling, retrieval-augmented generation, multi-agent workflows, and clinical prediction APIs.

At TechVision, I designed, trained, and shipped machine learning and AI models for client and internal projects, taking solutions from prototype through live production. I integrated LLM-powered features with Python, LLM APIs, tool/function calling, and RAG pipelines alongside the engineering team.

As a Research Collaborator at Government College Women University, I led a hybrid deep learning project from proposal and defense through deployment and expert review. My first-author Springer Nature paper presents a stacked BiLSTM clinical decision support model that achieved 96% accuracy and 100% sensitivity on an independent clinical test set.

I'm focused on safe, auditable healthcare AI: from MCP-wrapped XGBoost tools and guideline-grounded clinical agents to FastAPI prediction services and multi-document medical RAG assistants. I'm seeking a remote AI/LLM development role where I can ship practical systems and evaluate new agent tooling for real product use.