Living Lab

Proof, not promises.

Every system on this page runs on this website right now.

All systems operational·Last checked: 2026-04-07T15:42:14.573Z

AL-CHT-001Available

Answers visitor questions via floating chat widget.

FAQ match → Ollama → Claude API fallback

3-tier routing, all responses cached

AL-RSS-001/002/003Active

Curates 65 RSS sources into daily AI digests.

Fetch (30min/6hr/daily) → Deduplicate → Summarize with LLM

65 sources · 5 categories · Updated daily

AL-ANL-001/002Collecting

Cookieless visitor tracking and aggregation.

Pageview events → Daily aggregation → Dashboard

Zero cookies · Daily rotation hash · GDPR-native

AL-SEC-001Scheduled

Automated vulnerability and dependency scanning.

Daily scan at 03:00 UTC → Alert on findings

Daily at 03:00 UTC · ISO 27001

AL-SEO-002Scheduled

Detects broken internal and external links.

Weekly scan Sunday 04:00 UTC → Report

Weekly · Sundays 04:00 UTC

AL-MON-002Tracking

Tracks every LLM call with model, tokens, cost.

Continuous logging → Daily summary at 08:00 UTC

Every call logged · model + tokens + cost

n8n-first cost architecture

Every repeatable task is a workflow, not an LLM call. AI is the last resort — and when used, every token is tracked and cached.

User action → n8n (free) → needs AI?
  → No: done (free)
  → Yes: Ollama (nearly free) → Claude API (paid, tracked, cached)