Shradha
Kaushal.
Senior Data Scientist
Revenue Analytics
Senior Data Scientist
Revenue Analytics
I build ML systems that drive measurable business impact. Specialized in predictive analytics, optimization, experimentation, and product strategy.
ETL & Processing
ML & Optimization
Model Evaluation
A/B Testing
Deployment & Scale
Led end-to-end product development for ML-powered pricing vertical targeting senior housing sector. Drove product roadmap through A/B testing and user research, achieving enterprise adoption across 5 major clients and 40K+ properties.
Product Strategy • A/B Testing • Enterprise SaaS
Built Random Forest ensemble with SARIMA capturing non-linear event impacts. Recursive 365-day forecasting with rolling-window cross-validation enabled Extended Stay America expansion.
Random Forest • SARIMA • Feature Engineering • Rolling CV
Two-stage MIP solver for hospitality revenue maximization. Stage 1 optimizes room allocations. Stage 2 calculates shadow prices for automated booking controls.
MIP • CBC Solver • Python Multiprocessing • Shadow Pricing
Designed end-to-end experimentation infrastructure with causal inference frameworks. Implemented staggered rollouts in smaller markets to validate product adoption and pricing configurations. Analytics revealed 60% calibration skip rate, leading to UI redesign.
A/B Testing • Product Analytics • Staggered Rollouts
Built end-to-end AI agent for automating analytical Jira ticketing. Integrated AWS Bedrock and Claude 3.5/4.5 with live Jira APIs, performing real-time SQL diagnostics and synthesizing model context through MCP servers. Automated 100% of standard customer success workflows.
AWS Bedrock • Claude 4.5 • Jira API • MCP • R Shiny
Mathematical foundations of PCA, eigenvalues, and dimensionality reduction theory.
Medium • Linear Algebra
Practical guide with code examples, visualizations, and real-world datasets.
Analytics Vidhya • Python
Hidden Markov Models achieving 89.9% accuracy in pandemic spread prediction with geospatial analysis.
NECSI • Geospatial • HMM