Rishabh Rathi Email & Phone Number

Bengaluru, Karnataka, India

● Owned the vertical planning, prioritisation, and execution to drive data observability forward, integrating ML-based monitors with various latencies into a scalable alerting and RCA system. Reduced Alerting time by 75%, Reduced RCA time by 50%.● Achieved 95% data compression while preserving 85% of streaming data characteristics by creating a novel ML based Compression Algorithm, significantly accelerating development and release testing.● Prototyped and maintained the company’s driver behaviour scoring system, enabling accurate cross-driver comparisons and trip-based scoring for key clients.● End-to-End Ideated, Build, Launched Fleet Electrification Product, ran pilot with few customers unlocking $50K in revenue.

Bengaluru, Karnataka, India

● Designed and implemented a time series forecasting framework to detect anomalies in vehicle telemetry data and processedoutputs across multiple aggregation levels, flagging customer-specific data sources.● Reduced false positives by ~60%, significantly improving system efficiency and scalability for large-scale data monitoring.● Built a scalable MLOps framework for stateful streaming pipelines using Apache Spark, Databricks Delta Live Tables, DeltaLake optimising complex stateful features and reducing operational costs by 44% through native ML model implementations.● Improved EV efficiency modelling with regression techniques and optimising ML models for vehicle health assessments.

Thane, Maharashtra, India

● Conducted behavioural segmentation to analyse customer lifetime value (CLV) based on user interactions, accurately predicting lead potential and optimising sales strategies through actionable insights, reducing customer acquisition cost by 60%● Engineered and maintained scalable data pipelines to support marketing, sales, and customer retention, ensuring efficient data flow across business functions.● Developed a regression model using CRIF data, achieving 11% MAPE with CatBoost Regressor, identifying high-value customers for cross-sell strategies.● Built a Random Forest classification model with 97% precision (F1 score: 0.95) to target cold leads, increasing conversion rates by optimising lead scoring and pre-processing techniques.● Automated the "Refer and Earn" program, reducing manual workload by 90 man-hours per month, and collaborated with cross-functional teams to streamline program implementation and website dashboard updates.

Bachelor Of Science - Bsc, Data Science And Programming

Bachelor Of Technology - Btech, Computer Science With Specialization In Data Science