Economist, Data Scientist - Univeristy of Berkely
Ashok Bardhan is a consultant and advisor to a number of data analytics, finance and technology firms, public agencies, and to a large public pension fund. I have led and guided teams of data scientists, engineers and statisticians in solving a very wide range of business problems.I was Senior Economist at the Haas School of Business, UC Berkeley. My research and publications span a range of subjects including management issues of globalized innovation; impact of offshoring on jobs and firms; urban and regional development; housing, housing finance and the financial crisis; trade and technology linkages between US, China and India.I combine causal inference augmented AI/ML with econometrics for informed decision-making. I work on predictive solutions in finance(collections, forecasting, fraud prevention), retail/e-commerce (elasticities, customer segmentation, lead conversions, revenue optimization, churn), real estate (automated valuation, investor-scoring), in travel/tourism(dynamic real-time pricing, capacity utilization), on customer engagement (smart-routing, ranking-matching-pairing-nudging algorithms), in staffing/recruitment (efficient matching), in insurance (underwriting; claims analytics), in healthcare (locality-based social-economic-lifestyle determinants of health outcomes).Machine learning techniques do not lay great store by causal, structural models built on underlying behavioral relationships and motivations. Data analytics is full of domain-blind applications of ML to business problems, with no reference to causality, the underlying data generating process or microeconomic principles, resulting in non-interpretable, spurious and faulty results. Combining ML with Econometrics, i.e. statistics informed by insights from economics brings new perspectives to modern data science.One practical application where this approach is markedly superior is at working-out counterfactual -what-if-scenarios: e.g. predictions when a business or its competitors change prices.Previously, I was with the Reserve Bank of India.