Episode 01
What Is an AI Operations Platform?
Phase 0 — Planning
Now Live · 25-Episode Build Series
Build an
AI Operations Platform
for Your Hosting Business
Stop answering the same support tickets at 2AM. AIOP combines Claude AI, n8n, PostgreSQL, and HashiCorp Vault into a self-hosted, governed AI-NOC that diagnoses, auto-remediates, and escalates — with full audit logging.
0
Episodes
0%
Less Tickets
0
Tools Integrated
0:1
ROI at Scale
88-Second Auto-Remediation
L1 / L2 / L3 Governance
Claude AI + n8n + pgvector
Prometheus Fleet Monitoring
HashiCorp Vault Security
Slack ChatOps Interface
cPanel · WHMCS · Proxmox APIs
64:1 ROI at 1,000 Servers
What Is AIOP?
A complete, self-hosted AI Operations Platform that gives hosting providers, MSPs, and data centres an AI-assisted Network Operations Centre — governed, audited, and human-in-the-loop.
Claude AI Reasoning
Claude (Anthropic) analyses logs, diagnoses incidents, and recommends actions — grounded in your team's real runbooks via pgvector RAG. No hallucination.
n8n Workflow Automation
Self-hosted workflow engine wires every alert, agent action, Slack command, and approval gate together. Your data never leaves your servers.
Governed & Audited
Three-tier action model: L1 read-only, L2 auto-remediation, L3 requires human approval. Every action logged permanently in PostgreSQL.
Full Fleet Monitoring
Prometheus + Grafana + Loki across your entire server fleet. Deploy node_exporter to 1,000+ servers with a single bash script.
Hosting API Integration
Native integrations with cPanel/WHM, WHMCS, Proxmox, DirectAdmin, PagerDuty, and DNS providers via HashiCorp Vault-managed credentials.
Slack ChatOps
Your engineers type /aiop status and get an AI-generated fleet report in 5 seconds. Six slash commands. Six specialized AI agents.
The L1 / L2 / L3 Governance Model
AI-assisted means AI helps. You stay in control. Every action is typed, logged, and bounded by code — never by just a prompt.
Level 1 — Read Only
Diagnose & Inspect
Safe read operations. AI runs these autonomously 24/7 with zero risk to infrastructure.
- Check Apache / MySQL status
- Tail error logs
- Check SSL certificate expiry
- Query Prometheus metrics
- List DNS records
- Search OpenSearch logs
Level 2 — Safe Ops
Auto-Remediate
Reversible, routine operations. AI executes automatically — hardware security checked, audit-logged.
- Restart Apache / PHP-FPM
- Restart MySQL / Exim
- Force logrotate
- Renew SSL certificates
- Clear OPcache
- Block attacking IPs (CSF)
Level 3 — Human Approved
Critical Actions
Irreversible or high-impact operations. AI proposes, engineer approves via Slack before anything runs.
- Suspend hosting account
- Delete virtual machine
- Terminate hosting plan
- Wipe server disk
- Mass account operations
- Production configuration changes
The Technology Stack
10 production-grade open-source tools. Each chosen for a specific, non-overlapping purpose.
Claude AI
AI Reasoning
n8n
Orchestration
PostgreSQL
Data Store
pgvector
RAG / Semantic Search
Redis
Cache & Dedup
OpenSearch
Log Search
Prometheus
Metrics
Grafana
Dashboards
Vault
Secrets
Slack
ChatOps
bismarck@aiop-ctrl-01 ~/aiop $ docker compose up -d
✓ postgres started (pgvector ready)
✓ redis started (PONG)
✓ vault started (unsealed)
✓ n8n started (editor live)
✓ opensearch started (indices ready)
✓ prometheus started (47 targets UP)
✓ grafana started (fleet dashboard live)
7 services healthy · AIOP ready ⚡
bismarck@aiop-ctrl-01 ~/aiop $ /aiop diagnose web01.prod
diagnosis: PHP-FPM worker pool exhaustion
confidence: 94% · action: restart_php_fpm · level: 2
⚡ Auto-executed · Recovery: 88 seconds · Audit logged
The 25-Episode Build Series
Every episode builds something real. Follow along from a blank Ubuntu server to a production AI Operations Platform — step by step, command by command.
Share the Revolution
25 Google-branded social media mockups ready for Instagram, LinkedIn, WhatsApp, and Facebook Stories.
25 unique social media cards · Google colours · Portrait of Bismarck Wanjala · Ready to publish
Proven ROI for Hosting Providers
The AIOP platform delivers measurable, auditable business outcomes from day one. Every metric is tracked automatically through Prometheus and PostgreSQL.
AIOP Dashboard · Live
● LIVE
Built for Hosting Infrastructure Teams
Whether you run 10 servers or 10,000, AIOP scales with your business — and the build series shows you exactly how.
Web Hosting Providers
Manage cPanel, DirectAdmin, and Plesk fleets with AI-powered L1 and L2 support. Reduce ticket volume by 80% while maintaining full audit compliance.
Learn more →Managed Service Providers
Scale your MSP operations without scaling headcount. AIOP's six specialist AI agents handle website, database, email, security, monitoring, and billing.
Learn more →Data Centres
Fleet-wide Prometheus monitoring, predictive maintenance that warns 48 hours before failures, and self-healing playbooks that run at 3AM.
Learn more →DevOps Engineers
A real, production-grade build series. Every command shown live. Every config file available. pgvector RAG, FastAPI governance, Vault secrets — all from scratch.
Watch the series →NOC Managers
Full observability dashboard, AI-generated weekly health reports, MTTR tracking, and automation rate KPIs — everything your management team needs.
Learn more →AI-Forward Startups
Build AI-native hosting from day one. The same architecture that enables 99.99% SLA commitments without proportional support team growth.
Get started →"The hosting companies that build AI-native operations in 2025 and 2026 will compete on a fundamentally different cost structure than those who have not started by 2028. AIOP is not a demo or a toy project — it is a production-grade platform you can deploy on your own infrastructure, today, following this series step by step."
Frequently Asked Questions
What is an AI Operations Platform (AIOP)?
AIOP is a self-hosted system that combines Claude AI, n8n workflow automation, PostgreSQL with pgvector, Redis, OpenSearch, Prometheus, Grafana, HashiCorp Vault, and Slack to give hosting providers an AI-assisted Network Operations Centre. It diagnoses hosting incidents, auto-remediates Level 1 and Level 2 issues, and requires human approval for Level 3 critical actions — with a full audit trail.
What is Level 1, Level 2, and Level 3 support in hosting?
Level 1 support covers routine queries and read-only diagnostics — checking server status, reviewing logs, verifying SSL certificates. Level 2 technical support handles safe operational changes — restarting services, rotating logs, clearing caches — that are reversible and low-risk. Level 3 technical support covers critical, potentially irreversible actions — account suspension, VM deletion, production configuration changes — that always require explicit human approval in AIOP.
Can I really build AIOP from scratch by following the series?
Yes. Every single episode shows every command typed live on screen, with copy-paste commands available in the free AIOP Task List PDF. The series covers 25 episodes from a blank Ubuntu server all the way to a fully operational, production-hardened AI Operations Platform. Thousands of engineers across Africa and globally have followed along.
What hosting control panels does AIOP support?
AIOP natively integrates with cPanel/WHM, DirectAdmin, Proxmox (for VM management), and WHMCS (for billing automation). The modular architecture makes it straightforward to add additional control panels via the Tool Executor service covered in Episode 14.
How does AIOP handle web hosting automation at scale?
AIOP deploys Prometheus node_exporter across your entire server fleet with a single bash deployment script (Episode 16), then uses predictive Prometheus queries to identify failure trajectories 48 hours before they manifest (Episode 22). The n8n cluster mode covered in Episode 24 scales the automation layer to 1,000+ servers without bottlenecks.
Is AIOP suitable for small hosting providers?
Absolutely. The single-node deployment covered in Episodes 1-20 is perfect for providers managing 10 to 500 servers. The minimum spec Control Node (8 vCPU, 32GB RAM, 500GB SSD) costs roughly $40-80/month on most VPS providers, and the ROI on even 50 servers — through reduced L1 support tickets alone — is immediate.
Ready to Build Your AI Operations Platform?
Join hosting engineers across East Africa and beyond who are building AIOP. Free documentation, free series, free resources.
🌐 www.aiop.co.ke · ✉ build@aiop.co.ke · 📲 @mrbismarck88