HO16 — Final Capstone — AI Solution

Hands-on portfolio project · Week 6 · Peer-reviewed · module M19. Part of the ForgeMind AI — AI Productivity Essentials course.

This is the polished reference branch. Each sample shows a complete solution built two ways: the primary no-code build inside your Claude.ai subscription, and an optional advanced version that deploys the same idea as a live app.

No API key needed for the course. Everything in the primary path runs on your normal Claude.ai login (Projects, Cowork, Skills, MCP via Claude Desktop, Artifacts). The deployable app also runs in demo mode with no key — the API key is only for connecting it to live Claude.

Goal

Done when: you have built and shared an AI solution for one problem statement below, with a short write-up of the problem, your approach, the Claude.ai features you used, and the result.

What to ship

The built solution (a shared Claude Project + a published Artifact) + a story-driven write-up (problem → approach → tools used across the course → result) + a 3–5 min demo + a 'concepts used' checklist. (Optionally: a deployed URL, if you do the advanced build.)

Pick a problem statement

Open the matching sample folder — each has a Use it with your Claude.ai subscription section with numbered steps. Choose one — or bring your own (get it approved first):

1. AI Ops Command Center — operations run from memory; build one place to summarise your docs, answer grounded questions, and track key metrics.
2. Sales Intelligence Platform — no view of the pipeline; enrich and score leads, draft outreach, and see deal health at a glance.
3. Content Production Engine — research, drafting and tracking are disconnected; take a topic from brief to long-form to channel-ready variants.
4. Personal Finance Copilot — you never know where your money goes; upload a statement, categorise it, ask sourced questions, and see spending insights.
5. Customer Support Co-Pilot — support is slow and reactive; get cited answers from your docs, auto-triage tickets, and show a trends dashboard.

How to use this repo

1. Open index.html in a browser to see all five samples.
2. Pick a sample and follow its Use it with your Claude.ai subscription steps — no code, no terminal.
3. (Optional, advanced) Each sample also ships a runnable full-stack app; see its ## Optional — connect the live API & deploy (advanced) section.

---

HO16 · Peer-reviewed · ForgeMind AI Course · module M19 (Week 6)

💡 Use your Claude.ai Pro plan wisely

The Pro plan has a usage limit that resets every few hours. A few habits make it stretch — and keep a mistake from burning your whole session:

• Use the example prompt in each sample's README — it's already written and tested. Don't reinvent it.
• One clear prompt beats lots of vague back-and-forth. Say what you want, with an example, in a single message.
• Start a new chat when you switch tasks. Long chats re-read every earlier message and use up your limit faster.
• Don't paste big files over and over. Paste once, then refer back to it.
• If something works, keep it. Tweak it rather than regenerate from scratch.
• Using Claude Code or Cowork? This repo's CLAUDE.md makes Claude follow these same rules automatically, and SKILL.md is a reusable "token-wise" skill.

If you do hit the limit, it resets after a few hours — nothing you've saved is lost.

Run it locally

This is a full-stack AI solution (frontend + backend + database + AI calls). It runs on localhost — the GitHub Pages page is static and can't run the backend or the AI calls. To run it on your own machine:

# clone your copy of this repo, then:
cp .env.example .env      # database connection + your Claude/AI credentials
npm install
npm run dev               # starts the backend + frontend locally

Open the local URL it prints (e.g. http://localhost:3000). Each samples/sample-0X/ folder has its own README, architecture notes, and a concepts used checklist.

Deploying to a server

Getting the final solution to a live public URL is covered once in the course: See the Week-6 deploy walkthrough. Build and demo it locally first, then follow that one video to deploy.