Projects I’ve Built
Interactive prototypes born from real user research and deep product thinking.
10+ experiments | Agentic workflows, RAG, prompt engineering, full-stack AI products
As a Product Manager, I don’t just ideate - I research and understand pain points, validate with users, and build high-fidelity interactive prototypes that bring the full product vision to life. These are the personal projects and prototypes I’ve built in the last 12 months using Cursor, Vercel, Claude, LangChain/LangGraph, RAG pipelines, and no-code + code hybrids. Many started as class projects/ weekend experiments and turned into full demos I’ve tested with real users. Scroll, click “Try the Prototype” and see the prototypes.
As a Product Manager, I don’t just ideate - I research and understand pain points, validate with users, and build high-fidelity interactive prototypes that bring the full product vision to life. These are the personal projects and prototypes I’ve built in the last 12 months using Cursor, Vercel, Claude, LangChain/LangGraph, RAG pipelines, and no-code + code hybrids. Many started as class projects/ weekend experiments and turned into full demos I’ve tested with real users. Scroll, click “Try the Prototype” and see the prototypes.
NXT.ai - Feedback Intelligence Tool
Turning the daily flood of customer feedback into precise, ROI-ranked product decisions that actually move the needle.
As a Product Manager, I lived this frustration constantly: support tickets, surveys, app reviews, CRM notes, and transcripts pouring in from every direction, yet we still spent days manually sifting, debating, and guessing “what should we build next?” Industry data confirms the cost: over $100 billion in R&D wasted annually on fragmented feedback, 60% of shipped features going unused, and teams losing critical velocity to instinct-driven prioritization.
After extensive user research with PMs and founders, I designed NXT.ai: an AI-powered Feedback Intelligence Engine that unifies every scattered signal into a contextual “Product Memory Graph,” quantifies true business impact (revenue risk, churn probability, retention value), and delivers clear, ROI-weighted recommendations with closed-loop automation straight into Jira or Notion.
The live prototype I built and tested combines instant onboarding value and receive clean, deduplicated, urgency-ranked insights with prioritized actions in minutes with enterprise-scale intelligence: seamless CRM and support-system connections for continuous, adaptive foresight and real-time roadmap optimization.
As a Product Manager, I lived this frustration constantly: support tickets, surveys, app reviews, CRM notes, and transcripts pouring in from every direction, yet we still spent days manually sifting, debating, and guessing “what should we build next?” Industry data confirms the cost: over $100 billion in R&D wasted annually on fragmented feedback, 60% of shipped features going unused, and teams losing critical velocity to instinct-driven prioritization.
After extensive user research with PMs and founders, I designed NXT.ai: an AI-powered Feedback Intelligence Engine that unifies every scattered signal into a contextual “Product Memory Graph,” quantifies true business impact (revenue risk, churn probability, retention value), and delivers clear, ROI-weighted recommendations with closed-loop automation straight into Jira or Notion.
The live prototype I built and tested combines instant onboarding value and receive clean, deduplicated, urgency-ranked insights with prioritized actions in minutes with enterprise-scale intelligence: seamless CRM and support-system connections for continuous, adaptive foresight and real-time roadmap optimization.
TalentSphere - Early Career Recruiting Platform
Helping new graduates finally find roles, industries, and skills that actually fit - with credible, personalized guidance instead of generic AI noise.
I kept hearing the same stories from new graduates and early-career talent: “I don’t know which roles or industries are right for me”. “I have a vague idea of what I want, but no clue what skills I actually need”. They had ChatGPT and every other tool, yet still lacked a credible, trustworthy source that understood real market data, company requirements, and personalized fit.
That pain became the foundation. During a competitive hackathon, I designed and built TalentSphere as the complete product these graduates deserved: an intelligent, agentic platform that helps candidates discover the right roles and industries, maps the exact skills required, and then seamlessly connects them to opportunities through smart sourcing, nuanced matching, and human-sounding personalized outreach.
The prototype I shipped won “Most Innovative and Viable Product” because it delivered real clarity. A new grad pastes their background or interests, shares their unique story and the system instantly shows best-fit roles/industries, the precise skills they need to build, and then auto-generates tailored applications and warm recruiter messages that land replies.
This is the career-launch product I built for every new graduate who feels lost - turning vague ambition into confident, data-backed moves, while giving talent teams the fair shot at hiring them.
I kept hearing the same stories from new graduates and early-career talent: “I don’t know which roles or industries are right for me”. “I have a vague idea of what I want, but no clue what skills I actually need”. They had ChatGPT and every other tool, yet still lacked a credible, trustworthy source that understood real market data, company requirements, and personalized fit.
That pain became the foundation. During a competitive hackathon, I designed and built TalentSphere as the complete product these graduates deserved: an intelligent, agentic platform that helps candidates discover the right roles and industries, maps the exact skills required, and then seamlessly connects them to opportunities through smart sourcing, nuanced matching, and human-sounding personalized outreach.
The prototype I shipped won “Most Innovative and Viable Product” because it delivered real clarity. A new grad pastes their background or interests, shares their unique story and the system instantly shows best-fit roles/industries, the precise skills they need to build, and then auto-generates tailored applications and warm recruiter messages that land replies.
This is the career-launch product I built for every new graduate who feels lost - turning vague ambition into confident, data-backed moves, while giving talent teams the fair shot at hiring them.
Rotten Tom-AI-toes - GenAI Workflow Analyzer
Turning uncertainty around AI-powered clinical workflows into actionable insight for healthcare teams.
Being in Seattle, I realized healthcare is one of the largest and fastest-growing industries locally, and AI adoption was accelerating, but there was a critical gap. Generative AI workflows from radiology report generation to patient triage were being developed without a standardized way to assess safety, reliability, or compliance. The stakes were high: patient safety, regulatory adherence, and organizational trust were all on the line.
That problem became the foundation. I designed Rotten Tom-AI-toes, an enterprise SaaS prototype that guides healthcare teams through a structured evaluation wizard, simulates deep AI analysis, and produces a Freshness Score across six clinical dimensions: accuracy, safety, fairness, compliance, explainability, and robustness. The platform also provides actionable recommendations, a leaderboard comparing 17+ foundation models by suitability, and trends tracking to monitor evaluation history over time.
The live prototype I built combines a real-time risk gauge, data-rich dashboards with radar charts, score rings, and tabbed navigation, and an enterprise-grade UI inspired by platforms like Datadog and Palantir Foundry. It delivers clarity and confidence in minutes without any backend, helping healthcare teams safely adopt AI while keeping compliance and patient safety front and center.
Being in Seattle, I realized healthcare is one of the largest and fastest-growing industries locally, and AI adoption was accelerating, but there was a critical gap. Generative AI workflows from radiology report generation to patient triage were being developed without a standardized way to assess safety, reliability, or compliance. The stakes were high: patient safety, regulatory adherence, and organizational trust were all on the line.
That problem became the foundation. I designed Rotten Tom-AI-toes, an enterprise SaaS prototype that guides healthcare teams through a structured evaluation wizard, simulates deep AI analysis, and produces a Freshness Score across six clinical dimensions: accuracy, safety, fairness, compliance, explainability, and robustness. The platform also provides actionable recommendations, a leaderboard comparing 17+ foundation models by suitability, and trends tracking to monitor evaluation history over time.
The live prototype I built combines a real-time risk gauge, data-rich dashboards with radar charts, score rings, and tabbed navigation, and an enterprise-grade UI inspired by platforms like Datadog and Palantir Foundry. It delivers clarity and confidence in minutes without any backend, helping healthcare teams safely adopt AI while keeping compliance and patient safety front and center.
CatalytIQ - Analytics & Growth Automation Platform
Turning overwhelming D2C brand dashboards into clear, actionable growth moves that actually ship revenue.
CatalytIQ - Analytics and Growth Automation Platform
Turning overwhelming D2C brand dashboards into clear, actionable growth moves that actually ship revenue.
As a Product Manager who has worked closely with small D2C brands, I watched founders and growth leads stare at dashboards every single day asking the same exhausting question: “What should I actually do next?” Data everywhere, but zero clarity on the highest-leverage levers, no time to build campaign briefs, and churn creeping up unnoticed.
I built CatalytIQ as the complete growth co-pilot I wish every D2C founder had: an AI platform that connects withe-commerce platforms in one click, then surfaces the biggest growth opportunities, auto-generates ready-to-run campaign briefs, predicts churn before it happens, and instantly creates personalized retention flows.
The working prototype I shipped does exactly that - powered by RAG over real transaction data + LangChain. The growth co-pilot every small D2C team actually needs.
As a Product Manager who has worked closely with small D2C brands, I watched founders and growth leads stare at dashboards every single day asking the same exhausting question: “What should I actually do next?” Data everywhere, but zero clarity on the highest-leverage levers, no time to build campaign briefs, and churn creeping up unnoticed.
I built CatalytIQ as the complete growth co-pilot I wish every D2C founder had: an AI platform that connects withe-commerce platforms in one click, then surfaces the biggest growth opportunities, auto-generates ready-to-run campaign briefs, predicts churn before it happens, and instantly creates personalized retention flows.
The working prototype I shipped does exactly that - powered by RAG over real transaction data + LangChain. The growth co-pilot every small D2C team actually needs.