# Style Map — BertMahoney.com

Last updated: 2026-05-25

## Site

- URL: https://bertmahoney.com/weblog
- Platform: WordPress
- Lane: professional / technical writing

## Samples reviewed across time

2024:

- https://bertmahoney.com/redesigning-my-website-a-journey-towards-improvement — May 26, 2024

2026:

- https://bertmahoney.com/the-hard-problem-no-one-is-solving-in-ai-identity-continuity — Mar 25, 2026
- https://bertmahoney.com/from-ssh-friction-to-a-private-llm-cloud-building-a-private-repeatable-ollama-ec2-setup — Apr 25, 2026
- https://bertmahoney.com/i-built-a-private-ai-cloud-my-data-is-mine — May 1, 2026
- https://bertmahoney.com/running-local-ai-on-a-2017-imac-what-actually-happened — May 15, 2026
- homepage/case-study positioning at https://bertmahoney.com

## Core editorial identity

BertMahoney.com is the professional/technical lane. It presents Bert as an AI systems architect, hands-on designer, developer, and builder who cares about real workflows, persistent agents, local-first/private infrastructure, autonomy, and trust.

The writing should be technical-public: accessible to smart non-specialists, credible to technical readers, and opinionated from actual building experience.

## Tonal center

- practical
- direct
- field-tested
- allergic to hype
- willing to say “this failed”
- values sovereignty, privacy, portability, and open source
- personal enough to show why the technical choice matters
- professional enough to be legible to clients/collaborators

## Sentence rhythm

Common moves:

- short punch sentences after technical explanation
- direct contradiction of hype
- “Here’s what actually happened” framing
- numbered/bulleted technical breakdowns
- specific failure details followed by the lesson

Examples of compatible rhythm:

```text
No sponsored results.
No “It worked great!” thumbnails.
Just what actually happened.
```

```text
That’s not a tool.
That’s a very slow pen pal.
```

## Structural habits

BertMahoney.com pieces often work best when structured as:

1. concrete problem
2. why existing solutions were unacceptable
3. constraints
4. build attempt
5. failures/friction
6. final architecture or result
7. broader principle

Good section types:

- The goal
- What failed
- What worked
- Why this matters
- What I’m doing next
- Architecture
- Lessons learned

## Recurring themes

- local-first AI
- private infrastructure
- not letting platforms hold data hostage
- open-source models/tools
- AI continuity and persistent agents
- memory as UX + systems architecture
- real-world friction
- subscription fatigue
- workflows over demos
- trust through constraints
- separation of local data and remote compute

## Strong phrases / positioning patterns

- “My data is mine.”
- “Build like it.”
- “Not magic. More like correct.”
- “The friction is real. It’s front-loaded. And it’s worth it.”
- “The hard problem is not data portability. It is continuity.”
- “The goal isn’t recall. The goal is the user feeling held in memory.”

Do not overuse these; treat them as evidence of stance.

## What to preserve

- technical specificity when the project is public/approved
- clear constraints-first thinking
- actual errors/failures
- architecture diagrams or plain-text architecture when useful
- anti-hype honesty
- autonomy/data ownership politics
- builder credibility

## What to avoid

- generic AI thinkpiece tone
- breathless future-of-AI language
- vague “unlock productivity” phrasing
- SaaS-blog polish
- pretending a failed test succeeded
- hiding friction to look more expert

## Confidentiality rule

For in-development software products, keep details at 100,000-foot level before launch.

Public writing may name technologies like Hermes, WordPress, Substack, GitHub, open-source tools, local agents, etc. Do not name unreleased products, repos, detailed feature sets, or market angles unless Bert explicitly approves.

## SEO direction

Good keyword clusters:

- local AI
- private AI cloud
- open source AI
- AI memory
- AI agents
- persistent AI agents
- AI identity continuity
- data ownership
- local-first AI workflow
- WordPress AI publishing workflow

## Current priority essay

AI memory / Open Brain piece.

Target framing:

- technical public
- pro open source
- anti-platform lock-in
- anti “your data held hostage by subscriptions”
- archive ≠ memory
- memory should be portable, inspectable, provenance-backed, and governed by the person it describes