ai-fabrics ols dwdm
AI Fabrics × Open Line Systems — The Political Fight Coming in 2027

AI Fabrics × Open Line Systems — The Political Fight Coming in 2027

Rene Fichtmueller / 2026-06-24 / ~3 min read
LLM Journey AI from Zero Security MAGATAMA ShieldX AI Security Prompt Injection Defense Data Privacy
MAGATAMA, ShieldX, and the Hard Boundary Around AI

MAGATAMA, ShieldX, and the Hard Boundary Around AI

Discover the importance of hard boundaries in AI security with MAGATAMA and ShieldX. Learn how to prevent prompt injection and protect sensitive information.

Rene Fichtmueller / 2026-06-23 / ~2 min read
ieee standards ethernet 1.6t ethernet-standards network-planning technology-trends
The IEEE 802.3 Roadmap, Read Honestly

The IEEE 802.3 Roadmap, Read Honestly

Explore the real timeline for IEEE 802.3 Ethernet standards beyond optimistic projections. Plan your network deployments with accurate information.

Rene Fichtmueller / 2026-06-23 / ~5 min read
LLM Journey AI from Zero Codex Software Development Verification Practices
Building Software With LLMs: Do Not Trust Blindly, Verify

Building Software With LLMs: Do Not Trust Blindly, Verify

Building software with LLMs requires careful verification of changes to maintain quality and control.

Rene Fichtmueller / 2026-06-22 / ~2 min read
cpo packaging hyperscalers co-packaged_optics yield_challenges qualification_processes
Why CPO Won't Ship at Volume Before 2028

Why CPO Won't Ship at Volume Before 2028

Explore the hurdles preventing co-packaged optics from reaching volume production before 2028. Understand yield issues, lengthy qualifications, and cautious customer behavior.

Rene Fichtmueller / 2026-06-22 / ~4 min read
LLM Journey AI from Zero Training LoRA AI Training Tips Model Development
Training Needs Cooldown

Training Needs Cooldown

After each training run, take a cooldown to evaluate your results. This helps avoid unnecessary training runs and ensures better outcomes.

Rene Fichtmueller / 2026-06-21 / ~2 min read
LLM Journey AI from Zero Local AI LoRA Training
Adapters: A Small Training Run, Not a New Brain

Adapters: A Small Training Run, Not a New Brain

Adapters and LoRA in plain language: not a completely new model, but a small learned behavior layer on top of one.

Rene Fichtmueller / 2026-06-20 / ~2 min read
LLM Journey AI from Zero LLM Gateway Software Development
Why I Started Thinking About an LLM Gateway

Why I Started Thinking About an LLM Gateway

An LLM gateway is a front door for AI work: one place for routing, safety checks, audit, fallback, and cost awareness.

Rene Fichtmueller / 2026-06-19 / ~2 min read
adhd conferences nog
ADHD Diary #010: Conference Brain Doesn't Scale

ADHD Diary #010: Conference Brain Doesn't Scale

I just got back from three days of a regional NOG meeting. By the end of day one I had fourteen good ideas. By the end of day three I

Rene Fichtmueller / 2026-06-19 / ~5 min read
LLM Journey AI from Zero RAG Local AI
RAG: Giving the Model Good Notes Before You Ask

RAG: Giving the Model Good Notes Before You Ask

RAG is one of the most useful AI ideas for beginners: search first, give the model the right notes, then ask it to answer.

Rene Fichtmueller / 2026-06-18 / ~2 min read
// projects & tools
// about the author
Rene Fichtmueller
Social Networker
20+ years building and breaking networks. Optical transport, peering, data center infrastructure. Writing about the things that go wrong at 2AM — and how to prevent them.