Riddhesh Parekh

I’m a Data Science master’s candidate at RMIT University, focused on building and deploying AI models that produce probability-scored outputs and actionable decision support. I enjoy turning advanced analytics into clear insights for stakeholders through dashboards and monitored thresholds.

In my Data Science Intern role at Spe•trum Tuition (InfinityVerse AI), I built a Class Monitoring Panel that translated student skill-mastery scores into interactive network graphs and proficiency heatmaps for educators. I integrated model-driven skill analytics into dashboards, configured threshold-based alerts for at-risk students, and contributed nine REST API endpoints plus improvements to guardian authentication using Python, FastAPI, MongoDB, and LangChain.

As a Technical Research Analyst at Junomoneta Finsol Pvt. Ltd., I generated probability scores for predictive signals across financial datasets and supported model training and evaluation with cross-validation using precision, recall, and F1-score. I also set up threshold breach alerts and a feedback loop for continuous retraining, including back-testing against new market data.

I continue to strengthen my practical modeling skills through projects like an Employee Attrition Prediction Model where I built and deployed a multi-label model to predict and rank attrition risk with probability scores and achieved 85%+ accuracy. I also work on NLP and network analytics—such as sentiment/topic analysis and community detection—to deliver data-driven insights from complex unstructured and relational data.