Servers in the foothills.
Systems built to run█
Micro Mountain is my owned-hardware stack in Hinton, Alberta: private AI, storage, automation, web apps, and self-hosted services running on real machines. One operator. One rack. No mystery platform underneath.
The proof is the production environment running on it.
Everything below runs on the same hardware behind Micro Mountain. This is the proof-of-work stack.
Four things this stack proves.
Not a generic hosting pitch. This is what the environment already does for real projects, internal tools, and daily operations.
Self-hosted apps
VMs, containers, reverse proxies, identity, databases, and public routes for apps that need to run outside a laptop.
Private AI
Local models for chat, search-assisted workflows, tool calls, summarisation, and internal agents without handing every prompt to a third-party SaaS.
Storage & backup
Fast local storage for active services, bulk storage for media and backups, and enough operational discipline to know where the data actually lives.
Retail systems
Cannabis ordering, supplier data, POS exports, inventory logic, spreadsheet cleanup, and workflow tooling built by someone who works inside the problem.
Real iron, written down honestly.
The point is not pretending this is hyperscale cloud. The point is knowing exactly what is running, where, and why.
Things that actually shipped.
Public apps, private tools, and systems built because the off-the-shelf answer was not good enough.
A web app for visualising family trees from The Sims 4 save files. It turns messy save data into households, genealogy, traits, careers, aspirations, timelines, and shareable family views.
Built as a real product workflow: import, parse, summarize, render, and share, with privacy and storage limits treated as part of the product instead of an afterthought.
An Open WebUI-style chat platform built to control the provider layer, model list, streaming responses, workspaces, prompt modes, and tool approvals.
The goal is a practical AI interface that can bridge local models, hosted APIs, n8n workflows, search, file context, and business tooling without turning into a black box.
A monitoring and notification workflow built around n8n, service checks, logs, and local AI summaries. It exists to reduce noise, not generate more dashboards.
The useful part is context: what changed, what matters, and what can wait until morning.
An ordering workflow for cannabis retail. Drop in Greenline reports and AGLC catalogue data, then build consolidated purchasing and transfer logic across multiple stores.
It handles sales history, inventory, supplier availability, case quantities, SKU mismatches, and export files so ordering becomes a review process instead of spreadsheet trench warfare.
Micro Mountain is not trying to look like hyperscale cloud. It is owned hardware, written-down systems, and one person who knows where the cables, containers, routes, and databases are.
If you want to see how I think, this stack is the evidence: build the thing, run the thing, break the thing, document the fix, then make it easier to operate next time.