Quan Nguyen

I build the data foundations, governance, knowledge graphs, and agentic AI that turn enterprise information into something teams can actually trust and act on.

Get in touch LinkedIn ↗

quan@resume ~ %

$ whoami

now: Manager, AI, Data & Engineering at Ernst & Young
focus: Governance, knowledge graphs, agentic AI
building: WYN Living, a home-furnishings brand
learning: MS Analytics at Georgia Tech
based: Washington, DC metro area

$ stats --short

0yrs in data & AI
0enterprise orgs
0certifications
0AI-built products

01 · About

Data management meets applied AI

I lead enterprise data management, governance, and modernization efforts, building the data quality, stewardship, metadata, and access control layers that power analytics and context-aware AI across large organizations. Since 2018 I have worked across consulting and industry, moving from analytics and BI delivery into enterprise data architecture and, most recently, applied AI. I am currently pursuing an MS in Analytics at Georgia Tech to go deeper on the technical side of that work.

Knowledge Graphs

Designed and shipped an ontology-driven knowledge graph platform on Neo4j and GraphRAG, with natural language exploration, graph visualization, and role-based access control.

Governance

Built data quality, stewardship, metadata, access control, and shared business definitions into a consistent knowledge layer for analytics and context-aware AI.

Architecture

End-to-end solution architecture connecting data, knowledge, and application layers, transitioning current-state structures into scalable, governed platforms.

Delivery & Enablement

Delivered Power BI and Tableau dashboards, defined business process design, and built lightweight AI-assisted tools and reusable patterns for teams to adopt directly.

02 · Experience & Education

Career at a glance

Education above the line, work below · hover an entry for detail

James Madison University BBA, Computer Information Systems 2018

George Mason University MBA, Business Analytics 2023

Georgia Tech MS, Analytics 2025 – Now

Accenture Analytics & Data Architect 2018 – 2022

Led cross-functional teams to deliver analytics, ETL, and business intelligence solutions, including data models, dashboards, and platform integrations for large client environments.
Designed MLOps architecture with machine learning engineers and data scientists to support multiple machine learning use cases.
Delivered data protection and access solutions, identified product gaps, and led feature enhancements that improved adoption and user engagement.

Ernst & Young Senior Consultant, AI, Data & Engineering 2022 – 2025

Built data strategy, implementation plans, and operating models for a large aviation organization to strengthen data management and support the data product lifecycle.
Defined data design approaches and modernization plans that aligned data products to business priorities and improved decision-making.
Built data pipelines into Starburst with embedded data quality checks and delivered Power BI and Tableau dashboards to improve reporting and insight generation.

Ernst & Young Manager, AI, Data & Engineering 2025 – Now

Led enterprise data strategy, governance, and modernization efforts to improve how data is owned, accessed, and used across business and technology teams, with roadmaps tied to analytics, AI, and agent-based use cases.
Built core data management capabilities across data quality, stewardship, metadata, access control, and shared business definitions to create a consistent knowledge layer for analytics and context-aware AI.
Designed and deployed an internal ontology-driven knowledge graph platform using Neo4j, GraphRAG, and agentic LLM workflows, with natural language exploration, graph visualization, role-based access control, and Azure-ready deployment.
Built lightweight applications and internal tools using AI-assisted development, including early agent workflows and reusable patterns for AI intake, governance, and monitoring.
Provided end-to-end solution architecture for AI-ready data platforms that connect data, knowledge, and application layers, helping transition current-state structures into scalable, governed mobility and aviation ecosystems.

WYN Living Co-Founder, Head of Product & Operations 2025 – Now

Built and launched WYN Living, a home-furnishings brand built around natural materials, artisan craftsmanship, and contemporary design, and grew the range past 20 SKUs.
Ran product commercialization end to end: supplier coordination in Vietnam, product selection and specs, pricing, landed-cost analysis, packaging, and U.S. market positioning.
Set the e-commerce and go-to-market approach across Shopify and Amazon, covering catalog structure, merchandising, product copy, brand standards, and the digital customer experience.
Designed the operating model for overseas sourcing and U.S. fulfillment, weighing container freight, tariffs, warehousing, 3PL partners, inventory economics, and shipping.
Shaped the brand identity and customer proposition: visual direction, packaging, sustainability messaging, photography standards, and AI-assisted content.
Used pricing and margin analysis to balance competitive positioning, customer value, logistics cost, and long-term profitability as the catalog scaled.

03 · Projects

AI-assisted builds

Selected projects where I designed and built AI-powered applications end to end, from architecture through cloud deployment.

Functional
- Retrieval-augmented generation over enterprise knowledge for natural language search and context
- Custom knowledge graph editor with a personal workspace so people can contribute knowledge directly
- Connected to the broader data platform and data governance and catalog tools for added context
- Ingests files and suggests additions to keep the knowledge graph up to date
Technical
- React
- Python API
- PostgreSQL
- Neo4j
- Azure
React frontend, Python API, and PostgreSQL for operational data, with Neo4j as the knowledge graph layer. Led the team through architecture and delivery.
Functional
- A planning workspace for ecommerce importers who source in Vietnam and warehouse in Atlanta, used to work out what to sell and at what price before any cash goes out
- Planning side covers competitor tracking, a price ladder per SKU, FOB and landed cost, a 3D dimension viewer, Google Trends lookups, price history, and Excel catalog import
- Operations side is a small ERP: product and supplier master with pricing in VND, purchase orders dated to their FX rate, an inventory and stock movement ledger, sales orders, and a finance area for expenses, FX, and P&L
- Costing runs at the shipment level, prices the catalog at 60, 65, and 70 percent gross margin, and shows upside, base, and downside cases with a margin waterfall and a sensitivity view
- Simulation tools for demand forecasting, cash flow, FX what ifs, and reorder timing
- The flow goes from candidate products to supplier quotes, competitor checks, cost and margin scenarios, an approve or reject decision, then PO, receiving, sales, and finally SKU profitability and when to reorder
Technical
- FastAPI
- PostgreSQL
- React
- TypeScript
- react-three-fiber
- Docker
- AWS Lightsail
FastAPI backend with async SQLAlchemy over Postgres, routers kept thin and the work pushed into a service layer. Email and password auth is custom, with lockout and admin approval. The AI agents sit in their own package and each has a plain fallback, so the whole thing still runs without an API key. It ships as a few Docker containers behind Caddy on a single Lightsail box. Pricing and landed cost live on the product itself so they can't drift, and the shipment economics are a separate schema since a shipment is still a guess at that point.
Functional
- Digital resume with a dark/light theme toggle, scroll-reveal animations, and a sticky sidebar nav
- A 6-week self-study PM roadmap (Foundations, Execution, Analytics, Product Sense, Strategy/Behavioral, Mock Interviews) sourced from Cracking the PM Career and Cracking the PM Interview, with per-week and overall progress tracking and deliverable checklists
- Interactive drills: flashcards, a RICE calculator, a metrics-tree builder, timed estimation drills, a STAR story builder and story bank, a product-design worksheet, and a mock-interview simulator, with voice-to-text input on estimation and mock-interview answers
- Global auto-tagged notes, select-to-highlight text with a dedicated review page, and a searchable framework glossary available from any page
- An AI Coach that gives on-demand feedback on STAR stories, design worksheets, metrics trees, estimation answers, and mock-interview answers, grounded in the books' named frameworks (STAR, CIRCLES, AARM, RICE, Kano, Goal-Signal-Metric, HEART, Pyramid Principle), and flags off-topic or gibberish input instead of fabricating praise
- Site-wide PIN gate for access control, with all progress, notes, highlights, and saved work stored locally in the browser per device
Technical
- Vanilla JS
- Web Speech API
- AWS Lambda
- API Gateway
- OpenAI API
- S3
- CloudFront
Plain HTML, CSS, and JavaScript with no framework or build step, mobile-first with a single desktop breakpoint, and theming through CSS custom properties plus prefers-color-scheme. The AI Coach runs behind a Node.js Lambda that proxies to OpenAI's Chat Completions API through an API Gateway HTTP API, with CORS locked to the production domain, rate limiting, and an IAM role scoped to basic execution only, so the API key never reaches the client. The static site is served from a private S3 bucket behind CloudFront with Origin Access Control, HTTPS via ACM, and a 403-to-200 rewrite for client-side routing.
Functional
- Users connect their own data sources, including multiple sources at once, for scanning
- Scans data quality across six standard dimensions: accuracy, completeness, consistency, timeliness, uniqueness, and validity
- AI-assisted profiling that surfaces likely root causes and suggested remediations, not just raw rule failures
- Dashboard for monitoring data quality daily across connected sources over time
Technical
- React
- FastAPI
- LLM-powered
- PostgreSQL
- Cloud
React frontend with a FastAPI backend, with an LLM embedded in the pipeline to help profile issues and suggest remediations rather than relying on static rules alone. Scan results and history are stored in PostgreSQL to power the daily monitoring dashboard.