Data science foundation · enterprise AI delivery
I turn ambiguous operational problems into practical data and AI systems, connecting stakeholder discovery, solution architecture, implementation, evaluation, and adoption.
Applied analytics · ML evaluation · enterprise retrieval · responsible deployment
Selected portfolio
Case studies that connect the business problem, data and architecture, implemented system, evaluation, limitations, and operational value.
Portfolio evidence · in motion
Three durable signals from the portfolio, animated as you move through the evidence.
Portfolio snapshot · scroll-driven illustration, not a live feed.
Capabilities
PythonSQLstatistical analysisexperimentationKPI designdimensional modelingBIgeospatial analysis
Evidence: Operations Control Tower, public-sector analyticsRegressionclassificationcalibrationfeature engineeringSHAPsubgroup analysisdrift and temporal validation
Evidence: healthcare risk, tabular benchmark, SKMRAGembeddingsvector retrievaldocument ingestionOCRFastAPIDockergrounding and citation evaluation
Evidence: PolyRAG and document intelligenceFairness evaluationmodel cardshuman reviewauditabilityrisk registersdeployment gatesgovernance artifacts
Evidence: healthcare fairness, Aegis, public-sector AIProblem framingstakeholder discoverysolution architecturerequirementsimplementation planningadoption and value translation
Evidence: public-sector transformation, client analytics, enterprise prototypesResearch & selected credentials
Publications
Certifications
Experience
Explore
New DelhiHome
One connected practice
Across contexts
Different operating environments sharpen the same instinct: understand the system before proposing the solution.
The constant
Data, decisions, and products designed for the world people actually work in.
