Herman Green III Logo
Home Resume About Contact

Beacon Systems

Helping users move from discovering AI resources to applying them.

Project Beacon home screen showing the AI resource discovery landing page with featured resources, category navigation, and top-level layout
Industry
AI Tools / Productivity
Role
UX / Product Designer
Focus
Discovery, Information Architecture, Product Strategy
Timeline
Mar 2026 – Present

Built as a live MVP using Lovable, Supabase, Claude, and ChatGPT.

The short version

Problem
AI resources on LinkedIn were useful, but hard to reliably discover, organize, evaluate, and return to with enough context.
My Role
Solo UX and product designer across research, IA, visual design, QA, and product iteration.
What I Did
Designed a searchable resource library, category filters, resource modals, and implementation guides that connect resources to practical use.
Outcome
A working MVP that turns scattered creator resources into a curated library with clearer discovery, context, and workflow guidance.
Problem

Useful AI resources were scattered across a system not designed for implementation.

LinkedIn had become a major place to find AI tools, prompts, and workflows. But the platform was built for content discovery, not for helping people collect, organize, compare, and apply resources over time.

Research

Two friction points shaped the product direction.

Discovering resources reliably

Access depended on feed timing, creator networks, comments, and search behavior.

  • Follow the right creators
  • Comment to unlock links
  • Catch posts before they disappeared
  • Know exactly what to search for

Knowing what to do next

Saving a resource did not explain how to turn it into action.

  • When to use it
  • How it fit into a larger workflow
  • What to do after opening the link
  • Why one resource mattered more than another
Key Insight

People do not just adopt a tool. They adopt a clearer direction that includes one.

Concept Proposal

A searchable AI resource library and guide system that helps users find relevant resources, understand their purpose, and move from discovery to practical use.

Layer 1

Resource Library

  • Centralized scattered AI resources into one browsable destination
  • Supported search, filtering, and category-based discovery
  • Preserved the library context while users opened resource details
  • Created the foundation for deciding which resources belonged in workflows
Beacon browse library showing the V1 resource listing concept with sidebar category navigation, full resource grid, and search
Layer 2

Implementation Guides

  • Sequenced selected resources into guided workflow paths
  • Explained what each resource does, what it supports, and how to use it
  • Added context for users who needed direction, not just another saved link
Implementation guide modal showing step-by-step workflow instructions for an AI resource, including what it does, what it supports, and how to integrate it into a practical workflow
Validation / QA

Manual QA and iterative review replaced formal usability testing in this MVP phase.

Beacon’s current validation approach reflects its stage: a manually curated MVP without a formal testing cohort. Rather than running moderated usability sessions, I used structured QA passes to identify and resolve issues across the library.

Link Audit

Reviewed all library resources for broken or redirected URLs. Flagged and updated entries that no longer resolved correctly.

Category Taxonomy Review

Verified that resource counts in the sidebar matched the actual filtered results for each category. Fixed mismatches found during review.

Resource Modal QA

Confirmed description, use-case, and guide content for accuracy and consistency across resource entries in the library.

Implementation Guide Template Review

Ensured step structure and formatting are consistent across all guides. Each guide covers what it is, what it supports, and how to use it.

Cross-Device Layout Pass

Validated the browse library and modal views across desktop and mobile breakpoints. Confirmed no layout breaks in key interactive states.

Formal usability research, creator attribution verification, and automated link-freshness monitoring are scoped for a future iteration. They are not claimed as complete in this version.

Final Screens

Project Beacon

Open live prototype →
01
Home

Discovery landing page showing featured resources and top-level navigation.

Beacon home screen showing the main AI resource discovery landing page with featured resources and primary navigation
02
Browse Library

Full resource listing with left sidebar category navigation and main content area.

Beacon browse library showing the full resource listing with left sidebar and main grid
03
Category Filter

Creative category active: filtered results, visible resource counts, and highlighted sidebar selection.

Beacon browse library with Creative category selected, showing filtered resources and active sidebar state
04
Resource Modal

Contextual information at the point of discovery: description, use case, and trust signals.

Resource detail modal showing description, use case, and a link to the implementation guide for a selected AI resource
05
Implementation Guides

Curated index of step-by-step workflow guides organized by resource type.

Implementation guides index page showing a listing of curated workflow guides organized by resource type
06
Guide Modal

Step-by-step workflow context: what the resource does, what it supports, and how to use it.

Implementation guide detail modal showing step-by-step instructions for using an AI resource in a practical workflow
Outcomes / Reflection

Beacon taught me that solving a discovery problem often reveals a deeper context problem. Honest constraints make better products than overclaimed features.

Discovery problems are often context problems

The core friction was not that users could not find AI resources. It was that they could not evaluate or act on them once found. Reframing from “where to find links” to “how to understand and use them” changed the entire product direction and led directly to the implementation guide system.

MVP scope requires honest constraints

Beacon’s first version works for a manually curated, QA-reviewed library. Scaling creator attribution, link freshness verification, and community contribution are open problems. Not solved ones. Designing within that constraint produced a more trustworthy and useful product than overclaiming would have.