Experience

I worked across more than 600 SKUs in a technical diesel-parts catalog. My responsibilities included pricing strategy, digital sales operations, and improving how products were presented across online channels. The work required accuracy because customers often searched by manufacturer, part number, engine application, and compatibility.

I learned that conversion improves when the information around a product answers the buyer's practical questions quickly. Pricing matters, but so do naming, media, organization, and confidence that the selected component is correct.

I developed an AI-assisted workflow to place Bosch diesel-engine product videos across the eBay storefront. That gave customers more context without requiring the team to update every listing manually. The broader work contributed to a 15% quarterly sales increase.

This role showed me how operational improvements become commercial results. A better workflow can improve both the customer's experience and the team's ability to maintain a large catalog.

I founded Tap'n It Tech around custom NFC and RFID products that connected a physical item with a digital destination. I handled product development, customer outreach, explanation, and delivery. The venture generated more than $3,000 in revenue.

Because I was responsible for the entire process, I learned quickly which parts of a technical explanation helped a buyer and which parts only added noise. I had to identify the use case, make the benefit concrete, respond to questions, and deliver something that worked as promised.

That experience is the clearest early proof that I enjoy selling technical products, especially when the sale depends on making an unfamiliar idea easy to understand.

I evaluate AI-generated text, images, video, and speech against structured rubrics. The work calls for close reading and consistent judgment across instruction-following, factual accuracy, visual quality, content fidelity, and overall performance.

I separate factual errors from missed requirements and presentation problems, then explain the difference clearly. That habit transfers well to technical selling: understand what was actually requested, test a claim against evidence, and give feedback that another person can use.

I work with clinical data from more than 300 patients to support diabetes-risk prediction. I use Python to clean and normalize the records so the dataset is more consistent and useful for predictive modeling.

The project has shown me that analysis starts before a model runs. Definitions, missing values, normalization, and source quality shape what the result can mean. I bring that same evidence-first mindset to research and product conversations.