Publications

Abhishekh, H. and Faizan, S. (2012)
Australian & New Zealand Journal of Psychiatry
John Cade (1912- 1980), pp.68-69

Faizan, S., Raveesh, B., Ravindra, L. and Sharath,K. (2012)
BMC Proceedings
Pathways to psychiatric care in South India and their socio-demographic and attitudinal correlates.

Faizan, S., Raveesh, B,Anjali, V., Lakshmanagowda Sujatha,R. and Sharath,K. (2012)
BMC Proceedings
The attitude of non-psychiatry doctors to psychiatry and its correlates in Mysore, South India.

Poster presentations

 

The RAG Enhanced Presentation Platform (REPP) is an innovative web-based application designed to streamline the creation and delivery of academic and professional presentations. This platform arose out of my work for United Way Greater Toronto, an NGO dedicated to poverty alleviation in the Greater Toronto Area. My work was carried out under the supervision and guidance of Prof. Dr. Jay Qi. By integrating Retrieval-Augmented Generation (RAG) capabilities powered by OpenAI’s GPT-3.5-Turbo API, the platform enables users to construct interactive, AI-supported reports and presentations that cater to the needs of diverse audiences, including educators, students, and stakeholders. REPP provides a unified, responsive interface for querying structured data, generating insights, and presenting information with advanced AI support. It offers customizable features that allow users to personalize the content, style, and design of their presentations. Its intuitive interface facilitates efficient content management, including the upload of datasets and dashboards in formats such as CSV and Power BI/Tableau, and integrates MySQL and Firebase for secure database operations and user authentication. The RAG-based query assistant enables real-time analysis, allowing users to interact dynamically with uploaded documents to extract trends, recommendations, and insights, thereby enhancing decision-making processes and academic evaluations. The platform employs FAISS as a vector database to support efficient document retrieval, ensuring semantic relevance through OpenAI embedding models. To validate its efficacy, REPP underwent evaluation using the Stanford SQuAD dataset, with performance metrics including semantic similarity and BLEU scores used to gauge its accuracy in generating precise and relevant responses. Future advancements include the expansion of REPP into mobile applications for Android and iOS, integration with cloud platforms like AWS or Google Cloud, and enhanced personalization options for user experiences. By addressing challenges in traditional report generation and academic assessments, REPP has the potential to revolutionize content creation, making it accessible, interactive, and data-driven across educational and professional settings.

Do you want to collaborate on
projects?