Why Last-Touch Attribution Is Quietly Lying to Your Marketing Team
Last-touch attribution feels intuitive — give credit to the channel that closed the deal. But this logic systematically misfires, and your budget decisions are paying the price.
7+ years turning complex datasets into decisions leadership acts on — from predictive models and causal measurement to executive dashboards and data strategy.

I'm a Senior Data, Business Intelligence & Marketing Analytics professional with 7+ years of experience turning complex datasets into decisions that business leaders actually act on.
My technical work spans multi-touch attribution, prescriptive modeling, KPI reporting, mix modeling, incrementality testing, causal inference, and ML-powered propensity scoring. My BI work covers end-to-end dashboard ecosystems in Power BI and GA4, built for self-serve adoption across teams.
But what I'm best known for is the part that's hardest to automate: translating rigorous analysis into a clear story that earns trust in the room.
I'm currently targeting senior Marketing Science, Analytics, and BI roles where the expectation is not just to report the numbers — but to shape decisions.
Years in Analytics
Measurement Frameworks
ML Models Shipped
BI Ecosystems Delivered
2022 – Present
CoStar Group (LoopNet)
2019 – 2022
CoStar Group (LoopNet)
2020 – 2022
CoStar Group
2019
Chemours Company
2017 – 2018
Infosys Ltd.
2016 – 2017
Vuclip
University of North Carolina
Specialization in Data Science
University of Mumbai
ML & Predictive Analytics
ML propensity scoring engine on 185K-row multi-table B2B SaaS data. Hot-tier accounts convert at 80.6% vs. 26.5% baseline — a 3× lift. SHAP-explained Random Forest, AUC-PR optimised.
ML & Predictive Analytics
Kaplan-Meier + Cox PH survival analysis predicting who will churn and when — enabling retention interventions timed to individual risk windows.
Marketing Measurement Science
Causal geo-experiment across 40 US DMAs — DiD, synthetic control, and permutation testing recover 18% planted lift to within 1pp. iROAS 4×. Four methodological edge cases documented and fixed.
Data Storytelling & BI
End-to-end funnel from anonymous to authenticated journeys, packaged as a stakeholder narrative report that informed product and marketing strategy.
Financial Analytics
Segment-level CAC/LTV model surfacing which cohorts are structurally profitable vs. subsidised by blended averages.
Business Intelligence & Reporting
Executive dashboard suite replacing weekly ad hoc reporting — 60% reduction in analyst request volume, self-serve adoption across marketing and sales leadership.
Experimentation & Causal Inference
Reusable A/B testing framework with power calculation, sequential testing guardrails, and Bayesian updating — standardising experiment design and interpretation.
Customer & Market Intelligence
RFM + behavioural clustering across 10,000+ customers into 6 personas — defining the ICP that shaped territory design, outbound targeting, and roadmap priority.
18 projects across 7 analytics domains
ML & AI · Marketing · Product · Sales · BI & Finance · Experimentation · Customer
Last-touch attribution feels intuitive — give credit to the channel that closed the deal. But this logic systematically misfires, and your budget decisions are paying the price.
Most dashboards are built to impress, not to inform. Here is a framework for designing executive-facing analytics that drives decisions rather than gathering dust.