I build data and AI products that people keep using.

• One set of numbers instead of a hundred dashboards

  Amazon Ring, 2019 to 2022, BI engineer and project lead

  The situation

  Product, sales and leadership teams were working from more than 100 Looker dashboards that overlapped and often disagreed.

  What I did

  Talked to the people using them, agreed shared definitions for the core metrics, and rebuilt what mattered as 25 Tableau products. Tuned the Redshift SQL underneath so they refreshed faster.

  What changed

  Duplication down 35%, adoption up 20%, refresh times down 40%. A cohort view built along the way helped lift monthly retention 17% for one customer segment.

  What I took from it

  Most of the work was agreeing on definitions. The dashboards were the easy part.

• A roadmap four teams could deliver on

  Discover Financial Services, 2023 to 2024, product manager

  The situation

  The BI ecosystem for a Fortune 500 card network, spread across four business units and three offshore teams, each with its own priorities.

  What I did

  Owned the vision and roadmap and ran it as one backlog, using usage data and interviews to decide what came next. Automated the reporting work that was taking up analyst time.

  What changed

  Every iteration shipped on time. Delivery timelines came down 30% and report turnaround 60%.

  What I took from it

  A clear no, given early and in writing, kept more promises than a hopeful yes.

• A product built alone, end to end

  Skyspeak, 2024, independent

  The situation

  A weather app for Android with an assistant you can ask in plain language. No team, no handoffs.

  What I did

  Designed it, built it in Kotlin and Jetpack Compose, wired in the LLM assistant and published it on Google Play.

  What changed

  It's live, with more than 100 downloads. Small, and real.

  What I took from it

  Doing every role myself made me a more patient product manager.

• I talk to the people who use a report before I change it.
• I write decisions down, with the reason, on the day they're made.
• I would rather give a later date I can keep than an early one I can't.
• I check the numbers myself before they go in front of leadership.

Products I've designed, built and shipped on my own since 2024.

• Skyspeak (Weatherly) — Native Android Weather App

  Native Android weather app with a built-in AI assistant — instant on-device advice (umbrella, driving, hiking) with no network call needed, plus free-form natural-language Q&A backed by an LLM.

  Try itCode

• Multi-Agent Customer Support System

  Simulated enterprise support system with coordinated AI agents (router, technical, account) that communicate over a Redis message queue, each running as its own containerized microservice behind a Streamlit demo interface.

  Try itCode

• REAL ID Appointment Checker

  Automation script that monitors NJ MVC REAL ID appointment availability through scheduled headless-browser checks and sends real-time email alerts the moment a slot opens.

  Code

• Personalized Learning Recommender

  Agentic AI app that curates personalized learning resources based on your goals and interests, using Claude for LLM-based reranking with a Gradio interface and local caching.

  Code

• Restobot – AI Restaurant Assistant

  End-to-end AI product for restaurant discovery and decision support.

  Try itCode

• GenAI SQL Assistant

  Self-service GenAI analytics assistant that translates natural language into SQL, empowering non-technical users to query structured data and visualize insights in real time.

  Code

I started in 2011 as a data analyst writing ETL for Discover, then moved through BI development, Tableau architecture and Scrum Master work into product ownership. Nearly all of it was client work through one consulting firm, BitWise, where I stayed for thirteen years.

Certifications

Tableau Certified Data Analyst, Product Manager Certified, SAFe 5 Advanced Scrum Master, Certified Scrum Master, MCSA: SQL Server 2012

Education

B.E., Information Technology, Pimpri Chinchwad College of Engineering, Pune University (2006 – 2010)Applied Generative AI Bootcamp, Purdue University (Feb – May 2025)

Tools I use most

SQL, Tableau, Power BI, Snowflake, Python, LangChain, and the OpenAI and Claude APIs

If you're hiring for an analytics or AI product role, I'd be glad to talk.

New York City Metropolitan Area