Projects
Analytics, applied AI, and live demos
BrandSignal: BrandSignal is an AI visibility and Answer Engine Optimization platform that helps brands understand, measure, and improve how they appear across AI search experiences like ChatGPT, Grok, Gemini, and Perplexity. It evaluates visibility and recognition, competitor positioning, prompt coverage, citation readiness, and content gaps, then turns findings into a prioritized optimization roadmap.
SignalScope AI: SignalScope AI explores how marketers can move from detecting an emerging customer need to activating against it while the moment is still relevant. The platform monitors social conversations, search behavior, weather alerts, and travel disruptions, evaluates which moments are worth acting on, generates Meta and Google Search creative, and routes output through a human approval workflow with simulated performance feedback.
GrowthLab AI — Marketing Strategy Simulator: GrowthLab AI is a portfolio-grade growth-operator simulator where you run Lumora Skincare, a Series A DTC brand with a $250k monthly marketing budget, through 8 connected business weeks. Each week brings a market event, metric movement, war room diagnosis, seven-agent debate, experiment proposal, and a strategic decision that shapes what happens next. Built as a deployable product on deterministic mock data, with optional bring-your-own-key OpenAI for live decision interpretation.
CTV Ad Tech Lab: Programmatic Ads in a World Fueled by AI: YieldLab is an educational, full-stack simulator that opens the black box of programmatic Connected TV advertising. Structured as a guided three-act tour, from ecosystem basics to the ML pipeline to agentic AI, it lets you adjust campaign inputs, publisher floors, competition, auction type, and frequency caps and watch inventory forecasts, auction outcomes, ad-server waterfall logic, and an AI buying agent's recommendation recompute in real time. Built entirely on 100% synthetic data, with no real campaigns, exchange feeds, scraping, or PII.
Adaptive Gaming Experimentation: Most teams can run experiments; fewer can reliably decide what to run next. The Adaptive Experimentation Agent closes that loop: it reads past test results, validates data quality, estimates which variants worked, and recommends the next experiment, without auto-launching anything in production. Designed and shipped the end-to-end MVP as a Dell graduate capstone.
AI Marketing Strategy Agent: Built an AI-powered marketing consultant that audits a website's SEO health and automatically recommends paid media strategies, keyword opportunities, campaign structures, and landing page improvements — giving businesses agency-quality guidance in minutes, not weeks.
SignalTrail: SignalTrail is an AI-powered customer intelligence platform that transforms customer conversations into structured insights, helping teams uncover trends, prioritize feedback, and make better product decisions.
Smart Receipt Checker: Developed an intelligent receipt processing platform that automates the extraction, validation, and review of expense receipts. The solution leverages AI-powered OCR to capture key transaction details, applies business rules to identify discrepancies or missing information, and routes exceptions through a human-in-the-loop workflow for final verification.
LLM-Augmented Decision Support Tool: Built a deployable LLM-powered safety evaluation system for AI-generated educational content. Originally developed for the Data4Good competition as Regional Champions, this architecture frames factuality as a decision problem instead of a generation task using relevance-first gating and structured reasoning. The demo showcases how decision-oriented prompt engineering and hybrid ML approaches can create auditable and scalable AI safety layers.
Agentic AI Use Case Scorecard: Most teams struggle to identify where agentic AI creates real value. This interactive scorecard evaluates use cases across three dimensions, Agent Fit, Business Value, and Feasibility, using a weighted scoring model to determine whether to build, pilot, or deprioritize. It turns subjective AI decisions into structured, repeatable evaluations.
Socratic AI Tutor: An educational tool that uses large language models not to give answers, but to ask the right questions. Built on the Socratic method, it guides learners through problems step-by-step, fostering deeper understanding rather than rote memorization.
Exploring Seattle: A full-stack coding project featuring live weather forecasts, curated restaurant recommendations, outdoor activities, and local events around Seattle, all in one interactive experience.
AI Agents in Financial Advisory: As a team capstone project, we analyzed how agentic AI systems can transform financial advisory by moving beyond static automation into goal-driven, multi-agent decision workflows. Using a technology–market–society lens, we evaluated where AI agents create real value today, where risks emerge, and how firms can adopt them responsibly at scale.
Winery Consumer Analytics: Problem: Revenue concentrated in a small group of loyal customers while many disengage early. Approach: Used RFM segmentation and churn analysis on transaction-level data. Key Insight: Churn is predictable and front-loaded after early purchases. Impact: Enables segment-specific, lifecycle-aware marketing strategies that maximize ROI.
Coca-Cola Blak Relaunch Strategy: Modern functional beverage positioning for Millennials and Gen Z
Spotify Streaming Success Analysis: Data-driven insights into streaming performance patterns and audio features
GrowthLab AI — Marketing Strategy Simulator
CTV Ad Tech Lab: Programmatic Ads in a World Fueled by AI
Adaptive Gaming Experimentation
LLM-Augmented Decision Support Tool
AI Agents in Financial Advisory
Coca-Cola Blak Relaunch Strategy