Read the screenplay: FANNIEGATE — $7 trillion. 17 years. The biggest fraud in American capital markets.

Deep Dive

What Actually Worked

Three workflows turned Claude Code from a novelty into a genuine force multiplier. Each one changed how I think about what a solo developer can accomplish.

94% Goal Achievement Rate

18
Fully Achieved
37%
28
Mostly Achieved
57%
2
Partially Achieved
4%
1
Unclear
2%
01

Mass Content Generation at Scale

175+ pages in a single session

175+
Pages in peak session
18
Sessions on web content
1,000+
Total routes on the site

The biggest unlock was using Claude Code to generate content at a scale that would take a solo developer weeks. In my peak session, I shipped over 175 fully-formed pages — each with SEO metadata, structured data, internal linking, sitemap entries, and search index wiring.

The pattern: I'd describe the page template once, give Claude a list of topics, and let it run. Each page came out production-ready — not boilerplate, but genuinely useful content with unique data, interactive calculators, and cross-links to related pages.

I orchestrated this through parallel "agent waves" — spawning 3-4 sub-agents simultaneously, each building a batch of pages, then committing and moving to the next wave. On a good day this looked like a one-person content team outputting the work of five.

Key Lesson: The secret isn't the AI writing ability — it's the wiring. Claude doesn't just write content, it creates the page, adds it to the sitemap, wires the search index, generates OG image metadata, and cross-links to related pages. Doing all that by hand for 175 pages would take days.
02

Enterprise Platform CI/CD Sprints

22+ PRs in an 8-hour sprint

22+
PRs in one sprint
12
Sessions on platform dev
8h
Sprint duration

Enterprise platform development is usually slow — complex codebases, strict CI pipelines, multi-org deployments, test coverage requirements. I used Claude Code to compress what would normally be weeks of work into single-day sprints.

My biggest sprint: 22+ pull requests and two package versions shipped in 8 hours. Claude handled the full workflow — reading existing code, making changes, updating test classes, creating PRs with descriptions, debugging CI failures, and re-submitting until green.

The key was letting Claude own the iteration loop. Instead of me debugging each CI failure, I'd point Claude at the error output and say "fix it." It would read the error, understand the root cause, make the fix, and re-run — sometimes going through 5-7 cycles to get a single PR green. Tedious for a human, trivial for an AI.

Key Lesson: AI-assisted development shines brightest on tedious iteration loops. The CI debug cycle — run tests, read error, fix code, re-run — is exactly the kind of repetitive, context-heavy work that Claude handles faster than any human.
03

Cinematic Client Demos Under Deadline

Theatrical presentations built in hours

8
Demo sessions
~3min
Per iteration cycle
100%
Shipped on time

When you're pitching enterprise clients, first impressions matter. I used Claude Code to build interactive, cinematic proposal websites — not slide decks, but fully functional web apps with splash intros, data visualizations, interactive demos, and embedded games.

The workflow: I'd describe the vibe I wanted ("think Top Gun meets enterprise software"), provide the client data, and let Claude build it. Then I'd iterate — "make the intro more dramatic," "add an arcade game here," "make this chart animate on scroll." Each iteration took minutes, not hours.

The tight deadlines made this even more valuable. When you need a polished demo by tomorrow morning, having an AI that can build, iterate, and ship in real-time is the difference between a generic PDF and a presentation that gets a standing ovation.

Key Lesson: Claude Code's real superpower for client work isn't code generation — it's iteration speed. Being able to say 'change the entire tone of this presentation' and have it done in 3 minutes means you can explore creative directions that would be too expensive to try manually.

What Claude Does Best

The insights analysis tagged which Claude capabilities contributed most to successful sessions. Multi-file changes dominated — the ability to edit 10+ files in a single action is where AI development pulls furthest ahead of manual coding.

Multi-file Changes
Edits spanning many files in one action
36
Good Debugging
Correctly diagnosing root causes
5
Proactive Help
Suggesting improvements unprompted
4
Good Explanations
Clear reasoning about decisions
2

The Pattern

The common thread across all three workflows: Claude Code excels when you give it a clear outcome and let it figure out the implementation. The more you try to micro-manage the how, the less leverage you get. Trust the process, steer the output, iterate fast.

Get Glen’s Updates

Investing insights, new tools, and whatever I’m building this week. Free. No spam.

Unsubscribe anytime. I respect your inbox more than Congress respects property rights.

More From This Series