Skip to content

Client AI Portfolio - Conversational Concept

Freelance UI/UX engagement: chat-focused, accessibility-first portfolio concept with guided prompts and an AI-ready chat box. Lo-fi in Balsamiq, hi-fi in Figma.

Executive Summary

  • Designed a chat-style portfolio so recruiters can ask questions (About, Experience, Availability) or skim compact project cards (short blurb + optional Demo/Repo).
  • Introduced an AI-ready chat box: works today with guided prompts + scripted answers, and can plug into an LLM later without UI changes.
  • Accessibility was first-class: high contrast, keyboard operation, visible focus, and reduced-motion support (AODA / WCAG 2.0 AA oriented).
Project at a glance
  • Role: Freelance UI/UX Designer; Accessibility Tester
  • Deliverables: Lo-fi wireframes (Balsamiq), hi-fi flows (Figma), chat UI model, a11y acceptance criteria, developer handoff notes
  • Accessibility: Designed to conform to WCAG 2.0 Level AA (AODA); validated keyboard/focus patterns and reduced-motion behavior
  • Timeline:
    • Wireframes (Balsamiq): 2025-09-03
    • Project hiatus: 2025-09-04 → present (awaiting client content/brand assets)
    • Hi-fi: TBD
    • Accessibility review: TBD
  • Links: N/A

Context & Objectives

Client context The client needs to communicate credibility in seconds—fast access to resume, a concise picture of experience, and easy contact—without deep reading.

Objectives
  • Faster answers: guided prompts for “About me,” “Tools,” and “Availability.”
  • Compact projects: small cards with a one-line summary and optional Demo/Repo link.
  • Accessibility by default: predictable headings/landmarks, visible focus, and non-color cues.

AI Chat Box

How it works today (no backend)
  • Prompt chips (About · Experience · Availability · Projects · Resume) populate the input with a suggested question.
  • Scripted answers render as chat bubbles with scannable subheads and links (resume, featured projects).
  • Free-text input is enabled; unknown queries fall back to a helpful menu (“You can ask about…”) instead of dead-ends.
LLM-ready later
  • Same interface; swap the answer source to an API.
  • Content model supports summarization: short bios, one-line project blurbs, tags, and canonical links for grounding.
  • Safety & clarity: an “About summaries” note explains that generated answers are derived from on-page content.
Accessibility patterns (chat)
  • Keyboard: Tab to focus input; Enter sends; Shift+Enter inserts a newline; Esc clears suggestion.
  • Announce replies: new assistant messages use aria-live="polite" for screen readers.
  • Landmarks: <main> contains both chat and the project list; chat header is labeled (“Ask about [user]”).
  • Focus: focus returns to the input after send; visible focus rings throughout.
  • Reduced motion: typing indicators and scroll-to-latest respect prefers-reduced-motion.

Wireframing & Prototyping

Wireframe v4 — chat-focused, guided discovery
  • Right-rail prompts reduce cognitive load and route to key answers quickly.
  • Dual entry paths: conversational prompt chips and standard nav (Projects/Resume) for familiar scanning.
  • Compact project cards: title + short blurb + Demo/Repo link—keeps the page light and scannable.
  • Accessible defaults: large targets, visible focus, logical tab order, reduced-motion awareness.

Outcome & Next Steps

  • Alignment on chat-first discovery with compact project cards and an AI-ready chat box for speed.
  • Next: prepare hi-fi prototype, finalize brand visuals, expand the component library, and hand off for development with the established WCAG 2.0 AA acceptance criteria.

Status: On hold pending client content. Design deliverables are production-ready; development can resume immediately after asset delivery.