Mariam Hashmi

At IQ Springs, I integrate LLM-powered features and Python machine-learning models into production applications, and build Go services for internal products. I deploy and maintain services with Docker and CI/CD.

For Pulse Band, my capstone project, I built an AI and IoT worker-safety platform connecting biometric wearables, an Android app, AI services, and a React and TypeScript dashboard. I developed its real-time monitoring dashboard and services for fatigue and heat-stress prediction, and led a six-person engineering team.

In a University of Wollongong in Dubai group project, I implemented full fine-tuning and LoRA adaptation of multilingual DistilBERT for Urdu-English sentiment classification. The full fine-tuning variant reached 0.81 F1, while LoRA cut trainable parameters by approximately 168x. I'm particularly drawn to enterprise AI, healthcare technology, and internal tools that help other teams work effectively.