AI Venture Studio
The Unnamed Roads is an AI venture studio. 9 live projects. Built by one founder and a stack of AI agents.
📎 Stack: n8n · OpenClaw · LiteLLM · Coolify · Hetzner
Projects running
09
More efficient with agents
10x
vs traditional team
Lines of code auto-generated
120k+
across all projects
Live projects
Each project publishes AI-generated content daily, tracks analytics via Umami, and notifies via Telegram. No manual editing. No CMS logins.
Sports AI
Modern hockey analytics consulting
Modern data stack + flexible engagement model delivering on-demand hockey analytics squads for teams that want measurable tactical advantages without full-time hires.
Data points available
30M+
Enterprise AI
Distributed agent infrastructure
Infrastructure layer that enables AI agents to operate inside enterprise systems—securely, locally, and at scale. On-prem, edge, and hybrid deployment.
Deployment model
Edge + on-prem
Marketing
Hyperlocal GTM loops
Physical mail plus referral mechanics that trigger when neighbors buy, spinning up viral loops for brick-and-mortar brands.
Networks created
15
Logistics AI
Print routing from webhook to doorstep
Two-sided print routing platform. Retailers send orders via webhook – providers accept, produce, and ship with full tracking synced back automatically.
Print connections
2
Data Infrastructure
Validated training data with audit trail
Training data you can trust — with proof. Multi-annotator validation, confidence scoring, and full EU AI Act documentation delivered within 72 hours.
Label accuracy
92%+
Sports AI
Proactive hockey analytics research
Making Hockey Data Human. Proactive pattern detection in NHL data delivering insights to scouts, coaches, and analysts before anyone else sees them.
Agent Analysts deployed
10
Retail Tech
Clean price tags. Done in 5 minutes.
No printer. No stress. Pick a design, enter your price — get professional price tags delivered straight to your store in five minutes.
Orders
500+
Venture studio
AI-native solo venture studio
One founder. Seven projects. AI agents doing the work. Self-hosted on Coolify and Hetzner for ~35 EUR/month – replacing what would normally take 8-12 people.
AI agents deployed
8
Portfolio
Data engineering & full-stack development
Data Engineer in Stockholm. Modern data stack, Python automation, and scalable web solutions. n8n, Streamlit, Databricks, data chat agents.
Experience
10+ years
The thesis
Building a fully automated company where AI agents handle every operational role — on a cost-optimized platform designed for maximum profitability with zero headcount.
Agents as the only employees
Every operational role — content, research, publishing, analytics, notifications — is handled by AI agents running on schedule. No team. No overhead.
Cost-optimized for profitability
Running 9 live projects with daily AI content for ~400 kr/month. Coolify + Hetzner replaces 2 300+ kr/month in managed services — maximizing margin from day one.
Fully automated publishing loops
Research → draft → quality score → publish → IndexNow → newsletter — end-to-end autonomous. One human sets strategy. Agents execute everything else.
Daily pipeline
23 n8n workflows run staggered across the day. Each site has its own schedule, model routing, and publish path. One Hetzner VPS. Zero daily intervention.
One workflow pulls Umami stats for all 8 sites and packages the previous day into a single Telegram message. Paperclip CFO agent reads the same data for weekly financial ops.
Umami · Minio · Paperclip · Telegram
Each site runs its own keyword research window. THA pulls live NHL standings from the API. EIK reads performance data from Minio and scores topics with Groq. THB uses NHL data directly — no separate research step.
NHL API · Minio · Groq · n8n cron
THA at 10:00, FIN at 10:30, TUR at 11:00, TAN at 11:30, TAF at 12:00. LiteLLM routes each prompt to Groq or Gemini 2.5 Flash depending on the site. FIN generates in Swedish. THB uses Gemini directly with live standings data.
LiteLLM · Groq · Gemini 2.5 Flash · n8n
7 sites auto-publish without review. EIK articles stop here — a Telegram message with inline Approve/Reject buttons goes to me. One tap publishes, one tap discards. No login, no dashboard.
Telegram bot · n8n webhook · OpenClaw
OpenClaw writes the markdown file and commits to the site's GitHub repo. Netlify or Vercel deploys in ~60 seconds. IndexNow pings Google and Bing with the new URL. Resend fires the newsletter. One Telegram message lands in the Publishes topic.
OpenClaw · GitHub · Netlify · IndexNow · Resend
Stack
Every OS phase has named tools so partners can see exactly how we sense, compose, deploy, and audit autonomous ventures.
Everything runs on one Hetzner VPS managed by Coolify. No managed databases, no SaaS overhead.
Coolify
Self-hosted PaaS that manages all containers, env vars, and deploys on the Hetzner VPS.
Hetzner
CX32 VPS in Frankfurt. ~150 kr/month. Runs everything — n8n, LiteLLM, Minio, Umami, OpenClaw, Qdrant.
Minio
S3-compatible object storage for draft files, keyword JSONs, and performance insight data.
n8n runs 23 active workflows. Every schedule, API call, conditional branch, and Telegram notification.
n8n
Self-hosted workflow engine. Handles keyword research, content generation, publishing, analytics, and error monitoring.
Telegram
All notifications and approvals arrive here. Inline buttons for approving EIK articles without opening a browser.
OpenClaw
Webhook publish server. Receives article payloads and commits markdown to GitHub — one HTTP call per publish.
Paperclip
AI agent platform. CEO agent runs weekly strategy heartbeat. CFO agent reads Umami + LiteLLM spend data and flags cost anomalies.
LiteLLM proxies all model calls. Gemini 2.5 Flash runs article generation for 6 sites. Groq handles keyword scoring. Costs tracked per site.
LiteLLM
Self-hosted LLM gateway. Routes to Groq, Gemini, or other providers. Tracks spend. Single endpoint for all models.
Groq
Fast inference for keyword scoring and EIK article generation. Llama models. Free tier covers daily usage.
Gemini
Gemini 2.5 Flash is the primary article generator for 6 sites — TUR, TAN, TAF, TPR, THA, THB. Large context window, routed via LiteLLM.
Qdrant
Self-hosted vector DB for semantic search and RAG experiments on article archives.
Umami tracks traffic across all 8 sites. Daily Telegram reports with week-over-week comparisons.
Umami
Privacy-first analytics. Self-hosted. Daily automated reports sent to Telegram with pageviews, referrers, and top articles.
IndexNow
Instant URL submission to Google and Bing after every article publish. No waiting for crawlers.
Listmonk
Self-hosted newsletter platform. Subscriber list synced with Resend for transactional delivery.
Numbers
Full monthly breakdown for running 8 sites with daily AI content publishing.
“One VPS running Coolify, n8n, OpenClaw, LiteLLM, Minio, Umami, and Qdrant. Everything self-hosted.”
“Gemini 2.5 Flash via LiteLLM handles article generation for 6 sites. Groq handles keyword scoring and EIK. Spend tracked per site.”
“theunnamedroads.com, thehockeyanalytics.com, thehockeybrain.com — plus Netlify/Vercel free tiers.”
What this is
An AI venture studio — one founder, 9 live projects, self-hosted AI infrastructure.
The Unnamed Roads is an AI venture studio — one person, nine live projects, and an AI agent pipeline that handles everything from keyword research to publishing.
The entire stack is self-hosted on a single Hetzner VPS managed by Coolify. n8n orchestrates ~23 workflows. LiteLLM routes LLM calls to Groq and Gemini. OpenClaw commits markdown to GitHub. Netlify and Vercel deploy. Total cost: ~400 kr/month.
This site is where I write about what I'm building, what breaks, and what it actually costs to run AI infrastructure at this scale. No fluff. Just systems.
Fully self-hosted
Coolify + Hetzner replaces 2 300+ kr/month in managed SaaS. One VPS runs everything.
Autonomous publishing
Content research, generation, and publishing run on cron schedules without manual steps.
Telegram-first ops
Every workflow completion sends a Telegram message. Approvals use inline buttons. No dashboards needed.
9
Live projects
400 kr
Monthly infra
23
n8n workflows
Say hello
If you're building something similar — self-hosted AI infra, multi-site content ops, n8n workflows — I'm happy to talk. No sales pitch, just notes from the field.
Also on Bluesky: @theunnamedroads.bsky.social