Sourav
Banerjee
Data Governance & Engineering Lead at Daikin · Houston, TX
At Daikin, a Fortune 500 manufacturer, I've spent 2+ years leading enterprise data strategy and MDM programs. Outside of work I design and build user-facing products and tell stories through photography. In addition, I've studied Japanese for four years, including a year in Tokyo.
- Established a single source of truth across 12 business domains (Product MDM)
- Eliminated $100K+ in projected spend with LLM-based product classification
- Grew Atlan into the central data catalog and metadata hub for 300+ users
Experience
Daikin
Houston, TX
Data Governance Lead
- Led the Product Master Data Management (MDM) program end to end, from data strategy, data mapping, and integration architecture across SAP, AWS, Snowflake, Semarchy, and Matillion to stakeholder rollout. Established a single source of truth across 12 business domains, cut data acquisition time by 70%, improved consistency by 80%, and trained 200+ users.
- Architected an LLM-integrated product classification and taxonomy system using Snowflake Cortex with 85% accuracy, delivering a validated product hierarchy as a core asset for the enterprise MDM program and eliminating $100K+ in projected spend.
- Built ETL pipelines (Snowflake, AWS Glue, Matillion) extracting data from AS400 and other legacy systems into Tableau/reporting layers, reducing latency for downstream reporting.
- Deployed and owned Atlan’s growth as the enterprise’s central data catalog and metadata hub for 300+ users, standardizing cross-stack lineage across Snowflake, SQL Server, Tableau, ThoughtSpot, SAP, and Matillion to improve data observability, knowledge, and consistency.
- Drove company-wide adoption of a new data governance policy and top-down operating model across 10 business units, defining ownership frameworks, stewardship models, and shared business definitions that changed how the business manages data.
- Defined team-wide AI coding standards and integrated Cortex Code into the Snowflake development workflow for a team of 5 (scaling to 20), and led sprint planning for a two-person data engineering team.
- Directed rollout of a natural-language analytics platform (ThoughtSpot), enabling self-service BI for upper management across 3 executive dashboards and reducing reliance on ad hoc data requests.
Chaat
Houston, TX
Front End Lead
- Owned the full frontend lifecycle for a food-discovery platform (user research, Figma prototyping, A/B testing) as UI/UX design lead and website design lead, defining the product's theming and design language.
- Directed two implementation cycles (Flutter → React), managing scope and delivery from requirements through final feature release.
Neudesic
Houston, TX
Software Engineering Intern
- Worked with senior engineers in Business Analysis, Cloud Engineering, and UX Design on Azure cloud solutions for enterprise clients.
- Applied full-stack and analytical toolsets (Python, SQL, JavaScript, Tableau) to turn raw datasets into business reporting while delivering sprint goals on an Agile team.
Projects
Gakuji
React · Python/FastAPI · Electron · Claude Code
- Led end-to-end SDLC for a Japanese-learning desktop app, migrating backend services from AWS Lambda/SQS to a local-first architecture and shipping automated versioning/auto-update pipelines.
- Leveraged Claude Code to design, rebuild, and polish the app. Applied the same AI-assisted workflow to Daikin's Snowflake development.
Moto
React · Vite · d3-force · Supabase · Vercel · Claude Code
- Graph-based visual Japanese dictionary: each word branches into its kanji, and each kanji into every other word that shares it.
- Bundles nearly all of JMdict (~228K words, 13K kanji), indexed in a Web Worker. Graph rendered on d3-force physics rendered as React SVG. Every graph is a shareable URL.
- Two-way auto-sync with Anki: saved words go out as cards and learning status comes back, with Supabase accounts for cross-device sync and a .tsv export for Safari and phones. Server-rendered pages for all jōyō kanji.
Coursework
Sharehouse
All-in-one roommate app for shared expenses, balances, shopping lists, and real-time chat, built on Firebase auth, CRUD, and snapshot listeners. Term project for Android Development course.
Kotlin · Firebase · MVVM
CityLens
Web app for finding and organizing photography locations, with Google Maps and Firestore. Human-Computer Interaction term project, prototyped in Figma.
React · Firebase · Figma
Salt Fighter
1v1 fighting game built by an 8-person team; I worked on the in-game camera, character collision, sound design, and original soundtrack. Term project for Game Programming Paradigms.
Unreal Engine 5 · C++ · Ableton Live 11
Spotify Popularity Analysis
Machine learning analysis of what drives Spotify popularity; 61% accuracy with cross-validated neural net classification. Term project for Machine Learning course.
Python (Scikit-Learn, Jupyter Notebook)
Skills, Education & Honors
Technical Skills
- Data & Cloud
- Snowflake · Snowflake Cortex · AWS · Matillion · SQL Server · SAP · Tableau · ThoughtSpot · dbt · Excel
- Governance & MDM
- Master Data Management · Atlan · Semarchy · ELT · Data Lineage · Data Quality · Metadata Management
- Programming
- Python · SQL · Java · Kotlin · Swift · Dart/Flutter · React.js · Firebase · Docker · Git · Azure DevOps · Claude Code · GitHub Copilot · Cortex Code
- Design & Creative
- Figma · Ableton · PowerPoint
Languages
English (Native) · Japanese (Intermediate) · Bengali (Conversational) · Hindi (Basic) · Spanish (Basic)
Certifications
AZ-900: Azure Fundamentals · July 2022
Education
The University of Texas at Austin
Austin, TX · 2024
B.S. Computer Science · GPA 3.82
Certificate in Japanese · 1-year exchange at Waseda University, Tokyo
HCI · Machine Learning · Artificial Intelligence · Mobile Development · Game Programming · Operating Systems · Software Engineering · Networks
Honors
- Certificate of Global Engagement, UT Austin · Tokyo 2023
- Outstanding Japanese Language Student · 2023–2024
Contact
- Email[email protected]
- LinkedInlinkedin.com/in/sourav-ban
- ResumeDownload PDF