Scout
Look for repeated problems, not novelty for its own sake.
SMALL PRODUCT LAB. PRACTICAL AI SYSTEMS.
A personal lab for practical AI builds.
Agent workflows, memory systems, and shipped proofs. Built in public, one experiment at a time.
SITE TREE
- inde3d100.xyz/
- services/
- services.html — practical AI for businesses in N. Greece (EN)
- services.el.html — Ελληνικά
- projects/
- ai-agents-for-normies.html — beginner-friendly agent education
- obsidian-agent-memory.html — local Markdown brain for agents
- agent-house.html — peer-to-peer multi-agent workflows
- rolepilot.html — B2B onboarding prototype
- rpg-game-companion.html — coming soon
- latin-grow-greece.html — field lab (external)
- kairos.html — spiral day planner
- writing/
- git-workflow-for-non-coders.html — June 16
- kairos.html — June 16
- handoff-automation.html — June 5
- how-to-create-agent-house-room.html — June 4
- agent-house-pattern.html — May 29
- manual-memory.html — May 29
- questops.html — May 28
- obsidian-brain.html — May 29
- index.html — all field notes
- utility/
- services/
SERVICES · NORTHERN GREECE
The lab is also a local practice.
The work on this site is public proof. It's also the capability local businesses in Xanthi, Komotini, and Alexandroupoli can hire — coaching, team training, strategy consulting, AEO, and speaking. Calm, proof-first, no agency hype.
7 public surfacesAI education, agent memory, workflow builds, RPG companion, Latin Grow field lab.
10 SEO articles liveProduction content system tested on a real service site.
Safety-first AI pathAgent workflows built around approval, not blind autonomy.
OPERATING LOOP
Scout → Build → Publish → Compound
Build
Ship a small proof that someone can actually use.
Publish
Put the artifact online, explain the use case, and invite feedback.
Compound
Keep what creates demand. Kill or archive what only sounds interesting.
HOW WORK IS JUDGED
No theater. Proof first.
The point is not to look like a large studio. The point is to keep a sharp loop from signal to artifact to measurement.
Principle 01
Build the smallest version that can be criticized by reality.
Ideas become clearer when they have a URL, a user path, a call to action, and a way to tell whether anyone cared.
Principle 02
Tools should reduce uncertainty.
If a product does not make a decision, workflow, or learning path clearer, it is probably decoration.
Principle 03
AI needs boundaries.
The best systems make approval points visible instead of pretending autonomy is always progress.