My approach is definitely a systematic approach, in which I test the features multiple times but also document everything while testing the features. I also establish clear metrics and goals and understands the feature's requirements.

Built and shipped an AI-powered natural language search at JPMorgan Chase, led user research, wrote PRDs, defined OKRs, ran sprints in Jira, and analyzed adoption and search success with Tableau.

Prototyped multiple AI features and UX flows in Figma, ran usability tests, iterated on prompts, and documented results in Confluence for product and engineering.

Led an AI journaling app concept, ForwardNotes, with mood analytics and emotion tagging, created the Figma prototype, presented to judges, and placed 2nd in a PM ideathon.

Automated ISS WAP log analysis at NASA with Python, scripted parsing and anomaly detection, and informed comms reliability decisions

Built internal Python and SQLite tooling that generated hardware recommendations using rule based logic and prompt assisted comparisons for 50 to 75 weekly inquiries

Daily use of LLMs for research synthesis, test case generation, copy variants, SQL drafting, and code review, plus prompt libraries and A/B testing for quality

Comfortable evaluating models and features using task success, precision and recall style checks, latency tracking, feedback loops, and analytics dashboards

I’m outgoing, reliable, and quick to learn. I enjoy engaging shoppers, explaining products clearly, and hitting goals. I’m comfortable setting up demos, tracking KPIs, and representing Beam with a friendly, professional vibe.