Make hard questions answerable — by building the right tools, asking the right things first, and reading what the data reveals.
Who he is:Kanishka is a senior data scientist at Meta Fintech, where he owns measurement and analytics for a payment processing engine handling $100B+ in annual transaction volume. Before Meta, he spent two years at Fractal Analytics and seven years at Deloitte Advisory, working across payments, fintech, and enterprise analytics. His work spans causal inference, large-scale experimentation, ML model evaluation, and cost analytics — applied to some of the hardest measurement problems in consumer and business payments. He's from Mumbai. He studied Electronics Engineering, then got his MS from NYU. He believes the most valuable questions are the ones nobody thought to ask.
Not a fan of corporate bios? My bio, as interpreted by —
What I specialise in:
At Meta Fintech, a measurement decision touches $100B+ in annual transaction volume and hundreds of millions of users. Getting it right demands the full stack: statistical rigor, product judgment, and the cross-functional trust to make data actually move decisions.
The domain: consumer and business payments — authorization optimization, large-scale experimentation, cost intelligence, causal inference, platform reliability, and financial risk. The methods: SQL, Python, causal models, and ML evaluation frameworks applied where the stakes are highest.
Twelve years across Meta, Fractal, and Deloitte. The constant: making complexity legible, and turning analysis into action.
Selected work:
AI Agents for Payments Analytics
Meta Fintech · LLMs · Agent Development · Eval Design · Payments Domain SME
Built 4 analytical agents and 3 reusable recipes to scale payments data science work — eliminating repetitive manual work across auth, cost, and broader metrics analysis. Developed reporting automations to streamline inputs and free capacity for higher-order work. Serve as domain SME and quality gatekeeper for agentic outputs: building eval frameworks, overriding where payments expertise demands it, and enforcing a strict standard — accurate, defensible, and bite-sized enough to act on immediately.
Omni — Payments Engine Migration & Experimentation at Scale
Led end-to-end measurement for Meta's core payment processing engine — $100B+ in annual TPV. Executed 40+ global A/B experiments to validate and ship Omni, the largest infrastructure change in Meta Fintech's history: migrating 99%+ of TPV while sustaining 99.99% reliability. Omni now powers 30+ products and 50+ payment methods globally.
Authorization Rate Strategy — Industry Lagging to Standard
Built the definitive model of payment health at Meta — quantifying causal relationships between auth rate, retry sequencing, revenue collectability, chargebacks, and leakage. Translated findings into OKRs and network strategies that moved Meta's payment success rate from below-industry benchmark to industry standard, contributing directly to revenue collectability at $100B+ TPV scale.
Transaction Cost Intelligence — From Black Box to Managed Lever
Architected Meta's fee-level transaction cost tracking system, bringing granular visibility to payment economics across the full processing stack. Identified material savings across routing optimization, data corrections, localization, and same-currency settlement. Built the analytical infrastructure for PSP and network negotiations — contributing to $20M+ in negotiated savings.
Led end-to-end measurement for core payment engine; executed 40+ global A/B experiments, migrating over $100B+ in annual TPV with 99.99% reliability
Created a unified payment health model quantifying key causal relationships (e.g., authorization rate, retry, chargebacks); used findings to set OKRs and network strategies that elevated payment reliability to match industry standard
Architected the fee-level transaction cost tracking system for granular visibility into payment economics; identified material cost savings via routing, optimization, data sharing, localization, and same-currency settlement
Developed benchmarking frameworks and AI-assisted scenario models for PSP/network negotiations, resulting in $20M+ negotiated savings and informing executive decisions on mix-shift and co-branded card strategy
Established 0-to-1 measurement for Meta Fintech's privacy initiatives, developing playbooks and a repeatable framework for tracking privacy progress in probabilistic contexts
AI-Native Analytics
Built AI-native analytics workflows and domain-specific agents for payment auth and cost reporting, reducing manual analytical effort and increasing decision velocity
Lead domain SME for agentic outputs, designing evaluation frameworks and enforcing standards for actionable insights
Fractal Analytics · New York, NY
Jan 2022 – Oct 2023
Senior Consultant
Orchestrated data-driven decisioning for GWS business productivity software, optimizing growth strategies for the $1B+ SMB segment
Developed marketing business intelligence — defined taxonomy, data architecture, and migrated dashboards to live reporting
Forecasted the SMB impact of Duet AI adoption and the deprecation of Google Workspace Legacy — informing executive go-forward planning at a pivotal product transition moment
Delivered insights (funnel diagnostics, revenue, channel/region ROI) to inform forecasting, planning, budgeting, and performance strategy
Recognized twice by client and Firm Leadership for excellent delivery and rapid insights; NPS assessments consistently elevated from satisfactory to great
Deloitte Advisory · New York, NY
Aug 2015 – Jan 2022
Senior Consultant — promoted from campus hire in under 10 months · 5 performance awards
Lead 5-member product analytics team to build in-house fintech software solutions; worked alongside a 15-member scrum team to deliver client roadmaps
Designed dashboards, scorecards, and ad hoc reports using SQL/Excel and BI tools focused on investor statistics and regulatory reporting
Directed 0-to-1 build of compliance covering 10+ filing types across 5+ large-scale client deployments
Education
New York University, School of Engineering · Brooklyn, NY
Sep 2013 – May 2015
MS Technology Management (top 3 of cohort)
Serendipitous transition from building tech via engineering to understanding tech through the lens of management, investors, and markets
KJ Somaiya Engineering, University of Mumbai · Mumbai
June 2009 – May 2013
BE Electronics (Distinction Honors)
Inspired by the electronic developments of the 2000s — led to a lifelong interest in neural networks, lighting, batteries, and UAVs
Key Projects
Meta — Core payment processing infrastructure measurements to support Ads, wearables, subscriptions, tokenization, stablecoin, and agentic commerce
Fractal — GSuite deprecation & early forecasts of Duet AI for SMB
