I built an executive ready headcount and payroll dashboard to track team growth and planning signals. I automated the pipeline from Salesforce into Tableau to keep metrics current with minimal manual work. I enabled leadership to monitor staffing trends and connect operational health to capacity and spend decisions.

I built product health analytics for Dolby XP sales enablement. I implemented usage tracking and built a Power BI dashboard to answer who is using the tool, what content drives engagement, and where drop off happens. I turned an unmeasured internal product into a measurable one and informed which stakeholder rooms to prioritize.

I created live voice of customer reporting for the Dolby.com top tasks survey. I rebuilt legacy visuals into a live pipeline using Hotjar to Zapier to Google Sheets to Looker Studio so results updated automatically. I delivered a self serve dashboard that clarified user intent and supported roadmap decisions for personalization and content strategy.

I prototyped personalization for Dolby.com article discovery. I built an article recommendation model using navigation behavior and content similarity then deployed placements to enable measurement of impact on engagement. I reached about 66 percent alignment with observed reading patterns in offline evaluation.

I trained a spam detection classifier on email content and sender signals. I built a lightweight model to separate spam from legitimate messages and validated performance on held out data. I demonstrated how basic NLP features can reduce inbox noise and improve focus on important communication.

I built a playful terminal style tool at a vibe coding workshop designed to ship something unconventional fast. I made the Emotions Compiler turn user input into intentionally bad recommendations to explore how UX tone changes product perception. I focused on speed iteration and creative systems thinking.

I designed a bulk add flow to reduce friction when saving multiple songs to a playlist. I created wireframes and a demo video to show the interaction model and edge cases. I focused on reducing taps while preserving exploration context across playlists and artist pages.

I conceptualized a feature for podcast listeners to bookmark a moment and attach a note without breaking listening flow. I validated the pain point through quick user conversations and translated it into a prototype that supports reflection and structured recall. I added a data layer for understanding what moments trigger engagement.

I designed a recently played pattern to solve a common discovery failure when shuffle moves on before the user catches a track name. I built a wireframe and demo that aligns with Spotify exploration goals and reduces frustration during passive listening.