https://github.com/hk-dev13/100hires-portfolio-setup
This repository documents the setup process for modern AI-assisted development tools as part of the portfolio project assignment.
https://github.com/hk-dev13/100hires-portfolio-setup
Last synced: 13 days ago
JSON representation
This repository documents the setup process for modern AI-assisted development tools as part of the portfolio project assignment.
- Host: GitHub
- URL: https://github.com/hk-dev13/100hires-portfolio-setup
- Owner: hk-dev13
- Created: 2026-06-26T18:43:33.000Z (22 days ago)
- Default Branch: main
- Last Pushed: 2026-07-01T12:20:09.000Z (17 days ago)
- Last Synced: 2026-07-01T12:26:03.537Z (17 days ago)
- Size: 954 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 100Hires Portfolio Project
## AI-Powered SEO Content Production Research
This repository documents the setup process for modern AI-assisted development tools (Stage 1) and a comprehensive research project on AI-powered SEO Content Production (Stage 2).
---
# Stage 1 — Development Environment Setup
## Objective
This section documents the setup and configuration of Cursor IDE and key AI extensions as part of the initial assignment deliverables.
## Installed Tools
### 1. Cursor IDE
Installed successfully and used as the primary local development environment.
### 2. Claude Code Extension
Installed and authenticated successfully.
### 3. Codex Extension
Installed and authenticated successfully.
## Steps Completed
* Created a public GitHub repository.
* Configured local repository and verified remote connection.
* Installed Cursor, Claude Code, and Codex extensions.
* Completed OAuth authentication flows.
* Documented system verification steps and challenges.
---
# Stage 2 — AI-Powered SEO Content Production
## Objective
To analyze modern Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO) strategies by auditing the publications, video lectures, and posts of 10 leading digital marketing practitioners.
This research identifies actionable frameworks for optimizing brand visibility in a world dominated by LLMs (ChatGPT, Gemini, Claude) and generative search features (Google AI Overviews / AI Mode).
## Project Statistics
Reviewers can quickly assess the scale and depth of this research project:
* **10** Genuinely high-signal SEO & AI practitioners audited.
* **30** Recent LinkedIn posts summarized and analyzed (3 per expert).
* **30** YouTube video transcripts extracted via API, structured, and audited (3 per expert).
* **10** Deep-dive case studies, newsletters, books, or Substack memos analyzed (1 per expert).
* **≈70** primary research sources.
* **80+** Structured markdown files created and cross-linked.
* **100%** Clean workspace (all temporary Python extraction scripts and raw files purged).
## Research Methodology
This research combined manual curation with AI-assisted analysis.
## Tools Used
- Cursor IDE
- Claude Code
- Codex
- Git
- GitHub
- youtube-transcript-api
- Python
### Step 1 — Expert Selection
I manually identified ten practitioners who actively publish practical insights on AI-powered SEO content production. The selection prioritized **operators and practitioners** who run live experiments and manage search campaigns for high-growth B2B and B2C brands, rather than general theoretical influencers.
#### Why These Experts Were Chosen:
* **Michael King & Garrett Sussman (iPullRank):** Technical pioneers in search engine algorithms, patent analysis, and creators of the "Relevance Engineering" framework.
* **Aleyda Solís (Orainti):** A world-renowned technical SEO consultant who builds automated AI content auditing workflows and frameworks.
* **Lily Ray (Amsive):** The industry authority on Google's Quality Rater Guidelines, E-E-A-T, and tracking the real-world impact of AI Overviews on brand visibility.
* **Cyrus Shepard (Zyppy):** A veteran investigator who conducts large-scale, evidence-backed tests on link attributes, semantic signals, and entity optimization.
* **Britney Muller (Orange Labs):** A data scientist and former Hugging Face/Moz AI representative who bridges technical machine learning workflows with marketing automation.
* **Gael Breton (Authority Hacker):** A practitioner who runs massive affiliate/authority site portfolios, testing AI content scaling and programmatic playbooks on live assets.
* **Ben Leonard (Ecom Brokers):** An e-commerce entrepreneur who studies how AI search visibility affects brand equity, customer acquisition, and exit valuations.
* **Kevin Indig (Growth Memo):** A growth advisor who analyses search engines as interconnected systems and pioneered research on the "Ghost Citation" problem.
* **Ross Hudgens (Siege Media):** An agency operator who ran extensive B2B site benchmarks correlating transactional "Versus" comparison libraries with AI search referrals.
### Step 2 — Source Collection
For each expert, I manually collected:
- Recent LinkedIn posts
- Relevant YouTube videos
- Additional materials such as blogs, newsletters, podcasts, and case studies
LinkedIn posts and additional resources were curated manually to ensure they were relevant to the research objective.
YouTube transcripts were collected using the **youtube-transcript-api** after selecting the most relevant videos for each expert.
### Step 3 — AI-Assisted Documentation
After collecting the research materials, I used AI coding agents (Claude Code and Codex) to help organize the information into a consistent documentation structure.
The AI assisted with:
- Structuring Markdown documents
- Summarizing long-form content
- Extracting recurring themes
- Organizing key takeaways
- Maintaining consistent formatting
### Step 4 — Manual Review & Verification
Every generated document was manually reviewed before being committed.
During the review process, I verified:
- Source accuracy
- Summary quality
- Transcript relevance
- Markdown formatting
- Repository organization
Only after manual verification were changes committed and pushed to GitHub.
### Workflow Summary
Expert Selection
→ Manual Source Collection
→ YouTube Transcript Extraction (youtube-transcript-api)
→ AI-Assisted Documentation
→ Manual Review
→ Git Commit & Push
## Repository Structure
The research is organized systematically within the [research/](./research/) folder:
* **[linkedin-posts/](./research/linkedin-posts/)**: Analysis of 3 recent social posts per expert, focusing on dates, links, and immediate takeaways.
* **[youtube-transcripts/](./research/youtube-transcripts/)**: Blockquote-formatted transcripts of lectures and interviews, detailed summaries, and research notes.
* **[other/](./research/other/)**: Case studies, newsletters, books, and Substack growth memos published by the experts.
* **[sources.md](./research/sources.md)**: A centralized directory mapping social handles, websites, and research channels for all 10 experts.
## Expert Directory & Research Index
| # | Expert | Primary Focus | LinkedIn Research | YouTube Research | Other Materials |
|---|---|---|---|---|---|
| 1 | **Michael King** | Relevance Engineering & Patents | [Analysis](./research/linkedin-posts/michael-king.md) | [Transcripts](./research/youtube-transcripts/michael-king/README.md) | [Materials](./research/other/michael-king.md) |
| 2 | **Aleyda Solís** | SEO Workflows & AI Integration | [Analysis](./research/linkedin-posts/aleyda-solis.md) | [Transcripts](./research/youtube-transcripts/aleyda-solis/README.md) | [Materials](./research/other/aleyda-solis.md) |
| 3 | **Lily Ray** | E-E-A-T, Google Quality Raters, & Spam | [Analysis](./research/linkedin-posts/lily-ray.md) | [Transcripts](./research/youtube-transcripts/lily-ray/README.md) | [Materials](./research/other/lily-ray.md) |
| 4 | **Cyrus Shepard** | First-Person Content & Brand Signals | [Analysis](./research/linkedin-posts/cyrus-shepard.md) | [Transcripts](./research/youtube-transcripts/cyrus-shepard/README.md) | [Materials](./research/other/cyrus-shepard.md) |
| 5 | **Britney Muller** | Technical AI Literacy & Model Risks | [Analysis](./research/linkedin-posts/britney-muller.md) | [Transcripts](./research/youtube-transcripts/britney-muller/README.md) | [Materials](./research/other/britney-muller.md) |
| 6 | **Gael Breton** | Programmatic SEO & Authority Sites | [Analysis](./research/linkedin-posts/gael-breton.md) | [Transcripts](./research/youtube-transcripts/gael-breton/README.md) | [Materials](./research/other/gael-breton.md) |
| 7 | **Ben Leonard** | E-commerce Exit Positioning & GEO | [Analysis](./research/linkedin-posts/ben-leonard.md) | [Transcripts](./research/youtube-transcripts/ben-leonard/README.md) | [Materials](./research/other/ben-leonard.md) |
| 8 | **Kevin Indig** | Growth Systems & Information Gain | [Analysis](./research/linkedin-posts/kevin-indig.md) | [Transcripts](./research/youtube-transcripts/kevin-indig/README.md) | [Materials](./research/other/kevin-indig.md) |
| 9 | **Ross Hudgens** | Transactional Content Templates & ROI | [Analysis](./research/linkedin-posts/ross-hudgens.md) | [Transcripts](./research/youtube-transcripts/ross-hudgens/README.md) | [Materials](./research/other/ross-hudgens.md) |
| 10 | **Garrett Sussman** | AI Search Personalization & UX | [Analysis](./research/linkedin-posts/garrett-sussman.md) | [Transcripts](./research/youtube-transcripts/garrett-sussman/README.md) | [Materials](./research/other/garrett-sussman.md) |
---
# Research Outcome
Rather than treating each expert independently, I compared their findings to identify recurring patterns.
Across all collected material, six major themes consistently emerged regarding the future of AI-powered SEO and Generative Engine Optimization.
These themes represent the synthesized outcome of the research rather than the opinion of any single author.
## Key Research Themes & Actionable Insights
Through a cross-analysis of all 10 experts, six core pillars of modern generative optimization have emerged:
### 1. The Death of Educational TOFU (Blog Decay)
Basic informational queries ("what is X", "how to Y") are increasingly intercepted by AI Overviews and chatbots, resulting in zero-click searches. As a result, informational blog traffic is falling.
* **Actionable Reframe:** Move editorial focus from simple definitions to **Bottom-of-Funnel (BOFU) transactional content** (e.g., "X vs Y comparison pages," "X alternatives"), which retain click-through value and drive highly qualified conversions.
### 2. "TOFU with a Twist" & B2B2C
If a brand must publish informational content, it should not be written in a neutral, encyclopedia style. Instead, write with the bot in mind (**B2B2C: Business to Bot to Consumer**).
* **Actionable Reframe:** Explicitly state how your product or service solves the problem inside the body of informational articles (e.g., *"You can automate this segmentation process using HubSpot's email marketing software"*). This prompts LLMs to cite and recommend your brand when summarizing the topic, rather than parsing your data anonymously.
### 3. The Ghost Citation and Click Decoupling
According to research, 62% of brand citations inside LLM responses are "ghost citations"—where a chatbot links to the source URL but completely omits the brand name in the generated text. Gemini tends to name brands but rarely links, while ChatGPT links frequently but rarely names brands.
* **Actionable Reframe:** Shift KPIs from traditional ranking tables to LLM share of voice (recommendation rate) and branded search volume. Since LLM citation behavior is decoupled from traffic, marketers must evaluate search value based on overall branded homepage click surges and brand lift surveys.
### 4. RAG Optimization (Chunking, Semantic Triples, & Freshness)
Retrieval-Augmented Generation (RAG) engines retrieve content by scoring text "chunks" stored in vector databases.
* **Actionable Reframe:**
* **Structure:** Format articles into standalone, skimable sections with clear H2/H3 headers.
* **Grammar:** Write in "semantic triples" (subject-predicate-object) to make sentences mathematically parseable for cosine similarity checks.
* **Data Flywheel:** Maintain content freshness by embedding proprietary statistical surveys into your page templates quarterly; this justifies date updates that LLMs prioritize.
### 5. Google AI Mode & Hyper-Personalization
Google's AI Mode (Gemini) and Personal Context integrations will pull user-specific data (Gmail inboxes, receipt scans, Photos, location, device) to customize search recommendations, making universal rank tracking obsolete.
* **Actionable Reframe:** Focus on building a strong digital PR footprint and brand presence. Marketers must optimize for task execution (since AI agents will buy tickets and products directly from search) and ensure e-commerce feeds, schemas, and policies are fully parseable by bots.
### 6. Surround Sound GEO (Reddit, YouTube, LinkedIn)
Search engines are scraping user-generated platforms to inject human perspectives into AI overviews.
* **Actionable Reframe:** Optimize video transcripts (YouTube SEO), personal social pages (LinkedIn), and online community discussions (Reddit). Brand authority is built off-site; if you are cited on these platforms, LLMs will pull you into their core citation seed sets.