Project Demo
Jem&fix Ecommerce Assistant
Conversational commerce
A conversational ecommerce concept developed for Jem & fix, a leading Danish low-price DIY retailer. I designed and built this chat assistant prototype to allow customers to search for home improvement products, get practical DIY building advice, calculate materials, and add items directly to their shopping cart in real time using natural language.
Other Projects
Historical Investment Analytics with AI Agents
An AI-powered agent interface built to fetch and analyze historical investment data. Operating with Gemini models, this agent interprets user investment queries, retrieves historical stock and financial charts, computes analytical trends, and synthesizes visual financial reports with automated explanations.
UX Product Adviser in real time
A real-time UX advisory prototype that inspects UI layouts and recommends usability fixes. Running entirely on a local Large Language Model, it analyzes screenshot inputs to detect cognitive load, formatting, and layout risks, instantly serving actionable suggestions to improve user flows and conversion.
Local LLM as cybersecurity in Chrome
An experimental Chrome extension designed to scan webpages for security risks, phishing traits, and tracker vulnerabilities. By running a local LLM in the browser background, it audits active page contexts offline, ensuring high security and maximum user privacy without sharing web data with external APIs.
Connect with me on LinkedIn
Follow along on LinkedIn for more updates, tips,
and insights on UX design and AI.