SAI PRAHARSHA

Engineered and deployed three critical user-centric interfaces (robot command UI, nurse/admin dashboard, patient checkup flow), integrating real-time data visualization and responsive control to enhance operational oversight.

Developed an advanced end-to-end handle detection and manipulation system utilizing Faster R-CNN and GRU, achieving a 92% accuracy rate and significantly reducing manual intervention by 70%.

Collaborated effectively with cross-functional teams (robotics, design, clinical operations) to iterate on features, implement robust analytics dashboards, and inform data-driven product decisions via usage insights and A/B testing.

Refined GPT-based responses to align with user intent, boosting chatbot accuracy and relevance by approximately 30% and significantly increasing user satisfaction across chatbot products.

Executed rigorous fact-checking and prompt engineering, restructuring responses using VS Code, Excel, and internal annotation platforms to ensure consistent tone and factual correctness.

Curated and annotated extensive datasets for NLP and computer-vision pipelines, achieving over 10% gains in model precision and enabling robust, production-ready AI systems.