Open to PM, DA, BA & Consulting roles — available now
I turn messy operations into things that ship.
Two years as a business analyst at HDFC Bank, an NYU MS in Management of Technology, and a habit of staying with a problem until the fix is in production.
The short version of how I got here and what I'm looking for next.
I started out in electronics engineering at Mumbai University and found, between embedded systems, instrumentation and electrical machines, the part I actually cared about wasn't the hardware — it was the question of how to best utilize data to fill gaps between requirements and results.
I'm looking for product management, consulting, business and data analytics roles where the job values my people skills alongside my technical aptitude.
After engineering I joined HDFC Bank, where I spent 2 years as a business analyst translating between operations teams who knew the problems and the IT team who could enable it. I authored requirements, led the migration of 11 legacy .NET apps to Microsoft PowerApps, and built Power BI dashboards that replaced static Excel ones. The work I'm proudest of is an internal AI chatbot: I identified the need, scoped and developed it, and the result reduced routine support queries by 40%.
I recently consulted for Catholic Charities of Denver, evaluating over 15 real estate AI platforms and prototyping Tableau dashboards for 31 housing properties. This reduced their manual monthly data-consolidation process.
I also spent my final year teaching at NYU Stern School of Business, first as a TA for Research for Consumer Insights (Marketing), then leading problem-solving sessions for Microeconomics as an Instructor. Facing a room and making a complex concept understandable is the same skill as getting a stakeholder to sign off on a requirement: you have to understand it, break it down into simpler terms, and know what your audience already believes.
02 — Experience
Where I've worked
~3 years across banking technology, nonprofit consulting, university teaching and research, and industrial operations.
New York UniversityAUG 2026 — PRESENT
Research Assistant — Research for Consumer Insights · New York, NY
NYU Stern School of BusinessSEP 2025 — MAY 2026
Teaching — Economics & Marketing · New York, NY
Ran Microeconomics recitations for 3 sections as an Adjunct Instructor (Spring 2026) Teaching Assistant for Research for Consumer Insights — Marketing (Fall 2025)
Adjunct Faculty InstructorContractJAN — MAY 2026
Led recitation sessions for three sections of Microeconomics with Algebra, driving student engagement across the semester.
Broke down complex microeconomic concepts through intuitive explanations and practical, real-world examples.
Supported students beyond lectures by grading and running exam reviews that connected theory to application.
Teaching AssistantPart-timeSEP — DEC 2025
Supported 50+ marketing students in Research for Consumer Insights, evaluating assignments and holding office hours.
Worked with Dr. Carol Pluzinski to clarify concepts and strengthen student understanding of the material.
Facilitated hands-on training in R, IBM SPSS Statistics, and Qualtrics — the same research stack I use in my own analysis work.
Catholic Charities of DenverNYU Capstone ProjectSEP 2025 — DEC 2025
AI Project Consultant · Denver, CO
Won executive buy-in for an AI chatbot and cut 3–5 days of monthly reporting delay with Tableau prototypes across 31 properties.
Secured executive buy-in for an AI chatbot by evaluating 15+ real-estate AI platforms and writing the product requirements that made the case.
Prototyped real-time Tableau dashboards for 31 housing properties, eliminating 3–5 days of monthly data-consolidation delay and surfacing occupancy risk as it happened.
Delivered a phased technology roadmap off the back of 16 user discovery sessions, mapping reporting gaps, defining the metrics that mattered, and prioritizing around compliance constraints.
New York UniversityFEB 2025 — JUN 2025
Student Engagement Ambassador · New York, NY
Raised over $4,000 individually for scholarships and emergency funds for incoming students through alumni outreach by phone, text, and email, tracked in the RNL Engage CRM.
HDFC BankJUL 2022 — JUN 2024
Deputy Manager — Business Analyst · Mumbai, India
Scoped 18 applications, moved 11 legacy .NET apps to PowerApps for 500+ users, and shipped a chatbot that cut IT tickets 40%.
Authored User System Requirement Documents for 18 applications, gathering and prioritizing input from 20+ stakeholders across operations and finance.
Led requirements and delivery for migrating 11 legacy .NET applications to Microsoft PowerApps — enabling mobile access and ending manual security patching for 500+ users.
Replaced static downloadable reports with interactive Power BI dashboards embedded directly in PowerApps, so insight lived where the work happened.
Built SQL and RPA solutions across Excel and Power Automate that cut invoicing overhead by 7 hours weekly for a 25-person team and automated the daily IT support log.
Defined requirements for an internal AI chatbot on Power Virtual Agents that reduced routine IT support queries by 40%.
BCCI (ACC Cements)JAN 2022 — MAR 2022
Engineering Intern · Mumbai, India
Analyzed 5 years of operational data to identify peak operating windows, informing schedule optimization across the maintenance of 50+ instrumentation and control systems.
03 — Selected work
Projects
Analytics, data engineering, and financial modeling — end to end, from raw source to something a decision-maker can actually use.
Featured project · Financial services
Complaint Operations Console
Built a two-view investigation console in Palantir Foundry with embedded AIP Analyst. Prepared 99,898 CFPB complaints across 657 companies and surfaced 1,060 records flagged for review. The Consumer Financial Protection Bureau (CFPB) data covers approximately six months of 2026 across the United States: credit cards, checking or savings accounts, and money transfer, virtual currency, or money services.
One fault produces forty alerts, and the two loudest are usually the false ones. I built a correlation engine over a 30-service trading stack that collapses the storm into incidents and ranks the likely cause against the dependency graph — then wrapped it in a shift simulator where a clock runs and every move costs minutes. Classification is entirely deterministic; the language model only writes the prose.
Modelled a US consumer-electronics launch for an early marketplace operator, benchmarked against seven retailers. A $90 order leaves $7.68 once the card fee, carrier, returns and advertising are paid — and the third-party model that looks asset-light loses $6.02 an order, needing a 13% take rate to break even against the 6% assumed. The recommendation went the other way: start first-party.
A claims-investigation workspace that makes the next step visible: find the missing evidence, check it belongs to the case, draft the appeal, and keep a human responsible for approval. Eight screens over 30 authored claim scenarios, pulling real clinical context from a FHIR R4 sandbox; editing an approved draft sends it back to review.
Hand-labelled 391 nutrition panels in 27 languages out of a 4.7M-product corpus, then read each one with four different prompts. Where all four agree the answer is 82% right; where they disagree, 27% — so half the fields can be auto-accepted and half sent to a reviewer.
A subscription product for curated friend-matching meetups — four compatible strangers matched on overlapping interests, with a two-venue evening itinerary planned for them. Built the invitation and RSVP flow: compatibility scoring, venue sequencing, membership tiers, and the trust signals (ID verification, host coordination) that a stranger-meetup product lives or dies on.
Joined CMS readmission-penalty data for 2,373 US hospitals to their audited financial statements. My hypothesis was wrong: performance is uncorrelated with financial health (r = 0.01, survives five sets of controls). But because the penalty is a flat rate, it consumes 2% of net income at a healthy hospital and 40% at a struggling one — a 14–18× gap from an identical rule.
Pulled 58 years of GDP, population, and energy-use data across 10 countries from the World Bank API, built the ETL pipeline in Python and SQL, loaded it into Snowflake, and visualized how development and energy consumption actually track against each other.
Cleaned a 200,000+ record dataset spanning 50 years and built an interactive Tableau dashboard that lets you move from global pattern down to a single incident — turning an unwieldy archive into something you can actually ask questions of.
Mapped 1,713 actively exploited vulnerabilities to the ISO 27001 Annex A control whose failure each one represents. Four of the ninety-three controls carry 85% of the portfolio — and the classification is validated against a model and a blind manual sample, not asserted.
Evaluated a potential second-location acquisition with a 5-year revenue and profitability forecast, then stress-tested it — scenario analysis across pricing strategy, cost structure, and demand assumptions to find where the deal stops working.
SEP 2024 — MAY 2026 · TANDON SCHOOL OF ENGINEERING · NEW YORK, NY
MS, Management of Technology
GPA 3.64 / 4.00
Coursework
Project ManagementData Visualization for Business IntelligenceTechnology StrategyFinancial Analysis for Tech ManagersOperations ManagementDesign Thinking for Creative Problem SolvingEconomics & StrategyOrganizational BehaviorMarketingSupply Chain EngineeringGlobal Innovation
KJ Somaiya College
AUG 2018 — MAY 2022 · MUMBAI, INDIA
B.Tech, Electronics Engineering
CGPA 7.57 / 10≈ 3.2 / 4.0 equivalent
Coursework
Data StructuresProgramming in CConsumer BehaviorManagement & Organizational BehaviorBasics of AccountingSales & Distribution ManagementMobile App Development (Flutter)Virtual Instrumentation & Industrial AutomationEmbedded System DesignWireless Sensor Networks & IoTDigital Signal ProcessingCommunication Skills
06 — Contact
Let's talk.
I'm actively looking for product management and data analytics roles. If you're hiring — or just want to compare notes on making dashboards people actually open — I'd like to hear from you.
