AI Infrastructure Assistant

Your infrastructure,
one question away.

Ask about server health, investigate incidents, get optimization tips, or check security posture — all in natural language. The AI Assistant knows every server, service, domain, and alert in your infrastructure. Smart model routing delivers fast answers for simple questions and deep analysis when you need it.

2

AI models

<1s

Simple queries

$0.001

Per simple query

Auto

Model routing

Example Questions

What can you ask?

The AI Assistant understands your entire infrastructure context. Ask anything from high-level status checks to deep technical analysis. Here are some examples to get you started.

monitoring

Status

"How is my infrastructure doing?"

Get a complete status overview with server health, active alerts, capacity warnings, and recent incidents. One question replaces checking five different dashboards.

priority_high

Triage

"Which servers need attention?"

Get a prioritized list of servers with active issues, sorted by severity. CPU spikes, disk pressure, failed services, expiring certificates — ranked by urgency so you fix what matters first.

troubleshoot

Root Cause

"Why is api.example.com slow?"

Root cause analysis with correlated metrics across the stack. The assistant checks server load, database queries, network latency, and recent changes to pinpoint the bottleneck.

shield

Security

"Are there security risks?"

CVE scan results, expiring certificates, failed SSH logins, open ports, and outdated packages — all summarized with severity ratings and remediation steps.

tune

Optimization

"Recommend optimizations"

AI-powered config tuning suggestions based on your actual usage patterns. Right-sizing recommendations, MySQL/PostgreSQL tuning, nginx optimization, and resource rebalancing across your fleet.

history

Timeline

"What happened last night?"

Incident timeline and resolution summary for any time period. See what alerts fired, which servers were affected, how issues were resolved, and whether any follow-up is needed.

Smart Model Routing

The right model for every question.

We automatically route your question to the right model — fast answers for simple questions, deep analysis when you need it. You never have to think about which model to use. The router classifies your intent and picks the optimal path.

bolt

Simple Questions → Haiku

Status checks, server lists, alert summaries, certificate expiry. Fast responses under 1 second at approximately $0.001 per query.

psychology

Complex Analysis → Sonnet

Root cause analysis, optimization recommendations, security audits, incident post-mortems. Deeper reasoning at approximately $0.02 per query.

Model Routing Example

"Which servers are online?"

Haiku
Intent: status check · ~0.3s · ~$0.001

"List active alerts"

Haiku
Intent: list query · ~0.4s · ~$0.001

"Why is db-prod-01 slow?"

Sonnet
Intent: root cause analysis · ~3s · ~$0.02

"Optimize MySQL on web-prod-02"

Sonnet
Intent: optimization · ~4s · ~$0.02

Plan Availability

AI Assistant by plan.

The AI Assistant is available on Starter plans and above. Query limits scale with your plan so you always have the capacity you need.

Free

$0/mo

block

Not included

Use AI Analysis with your own API key instead. The AI Assistant is available starting from the Starter plan.

Starter

$9/mo

check_circle

20 queries/day

Perfect for daily status checks and quick troubleshooting. Covers most individual developer and freelancer needs.

Pro

$29/mo

check_circle

100 queries/day

Designed for teams running production infrastructure. Enough capacity for multiple team members and complex analysis workflows.

Enterprise

Custom

all_inclusive

Unlimited

No query limits. Priority model access with dedicated throughput. Custom model configuration and SLA guarantees.

Ask your infrastructure anything.

Start with the free plan and upgrade to unlock the AI Assistant. Natural language queries, smart model routing, and deep infrastructure context — all built in.

$ curl -sSL https://install.hostatlas.app/install.sh | sudo bash -s -- --key=SERVER_KEY_