A JavaScript-based personalized AI system with dual-memory architecture, real-time conversation analysis, and secure chat isolation for intelligent, context-aware interactions.
I've built a new mobile AI assistant that brings chat, text and image generation, voice input and document analysis into one app.
It have Smart Chat, Text Creator, Image Create, Voice Input, plus PDF Scanner, Photo Analyze, Social Content and Prompt Ideas, filtered by category.
Almost everything AI can do, just in one place.
Designed the landing page for Smug — an AI support agent that resolves tickets, knows exactly when to escalate to a human, and has zero interest in pretending otherwise.
Most AI support tools go the humble, over-apologetic route. Smug does the opposite — the whole brand is built around a smug little cat mascot and copy like "Handled it. You're welcome." It's a genuinely fun brief: figuring out how to make confidence read as charming instead of obnoxious.
Bold electric blue and orange instead of the usual muted SaaS palette, because a product this confident shouldn't look shy about it. 💙
I'm building Halvkoppling for the @Lovable Challenge on @Contra. It is a booking chat for a driving school in Stockholm with one instructor. Emil teaches from 08:00 to 18:30 and reads messages at lunch and in the evening.
The clip shows the learner's side on a phone, recorded on the live app:
1. The learner sends "My test is on 12 Nov. Manual, Risk 1 done."
2. The chat fills in the test date, the manual car and Risk 1 from that message.
3. The chat asks for the rest with tappable answers: Risk 2, the pickup stop and the weekday times.
4. The chat offers two plans that fit before the test, each with its price and its first two lessons placed in Emil's free times.
5. The learner pages to Plan B and back to Plan A, where the button reads "Confirm · 10 400 kr".
From there the learner pays in Stripe Checkout, or taps Send to a parent and a parent pays from a plan link. The same booking runs on WhatsApp.