‘Lab’ is a collection of AI experiments, tools, agents, and systems I've built to solve real problems
Some work may be in progress.
Most job searches are reactive, you scroll when you have time, apply when something looks right, and lose track of what you've sent out and why. For someone with a targeted career strategy and a specific type of role to find, that approach is tedious and time-consuming.
This project is an agentic system built on Replit that automates the discovery and evaluation layer of a job search; scraping target company career pages on a schedule, pre-filtering roles against a set criteria, and scoring each opportunity against a live master profile built from resume data, LinkedIn history, portfolio work, and 200+ real application outcomes. Built for busy individuals on a targeted job search, rhe result is a system that gets sharper the more I use it, surfaces only the roles worth my attention, and writes directly into my Notion dashboard, so I spend my time applying, not searching.
Ethical & Responsible AI: Synthetic Personas
Before building an AI system that simulates human research participants, I wanted to understand where that crosses a line.
This self-initiated research, conducted as the foundation for Verizon's Gem-powered synthetic user system, examines the ethical boundaries of AI-generated proxies in research contexts: when they're valid, when they're misleading, and what guardrails responsible deployment requires. Most practitioners build first and ask these questions later. This project started with the questions.
Interactive Service Blueprint
Traditional service blueprints have a shelf life problem. Built in Figma, shared as a static file, and obsolete the moment something changes, most blueprints are outdated before the team finishes reading them.
This work exploration project reimagines the service blueprint as a living system rather than a static artifact, connecting directly to live qualitative and quantitative data sources so the blueprint updates as the experience does. The result is an end-to-end view of main customer journeys that product managers and experience designers can actually trust, use, and share.
Most career advice is too generic to act on or too prescriptive to fit how people actually grow.
This project applies backcasting & Richard Boyatzis' Intentional Change Theory, a research-backed framework for sustainable personal transformation, to the specific challenge of career navigation.
It translates a five-stage psychological model: ideal self, real self, learning agenda, experimentation, and trusted relationships, into a structured design process that feels native to how designers already think: discovery, synthesis, prototyping, and iteration, but applied inward.
Built for mid-career individuals at an inflection point, this tool provides a set of prompts and reflective exercises that move you to a grounded picture of who you're becoming and what steps actually get you there.
Project type
Personal project, Agentic AI
Tools
Replit, Claude, classic notebook
Project type
Self-initiated research
Tools
Figjam, white papers
Project type
Work project
Tools
VS Code, Quantum Metrics, Adobe Analytics, & other internal tools
Project type
Personal project
Tools
Lovable, Miro