Frequently Asked Questions

CloudAIPilot is an AI-powered cloud management platform that lets you provision servers, deploy apps, manage backups, monitor infrastructure, control team permissions, and track costs — all from one dashboard. The AI Pilot sits on top of all of it, letting you operate your cloud using plain English instead of CLI commands. It works across AWS, Google Cloud, Azure, and DigitalOcean.

CloudAIPilot is built for founders, indie developers, small engineering teams, and digital agencies who manage cloud infrastructure without a dedicated DevOps team. If you've ever wished for a faster, safer way to manage cloud resources — especially without deep command-line expertise — CloudAIPilot is designed for you.

No. The AI Pilot is designed so non-technical founders can manage cloud infrastructure through natural language. You describe what you want — "create a WordPress site on a 2GB server" — and the AI Pilot handles the technical steps, walks you through its plan, and waits for your approval. Developers will still find deep controls (SSH terminal, firewall rules, Compose stacks), but you don't need any of that to get started.

Yes. The Free plan is permanently free — no trial, no credit card, just sign up and start. Every paid plan (Solo, Studio, and Agency) comes with a 14-day free trial, no credit card required. If CloudAIPilot isn't right for you during that window, you won't be charged.

Sign up at app.cloudaipilot.com — no invite code needed. Once in, connect your first cloud account (AWS, GCP, Azure, or DigitalOcean), and your dashboard is live. From there you can provision a server, deploy a site, or ask the AI Pilot what to do next.

CloudAIPilot currently supports Amazon Web Services (AWS), Google Cloud (GCP), Microsoft Azure, and DigitalOcean. All four providers can be connected simultaneously and managed from the same dashboard. Additional providers are on the roadmap.

Yes. You can connect as many cloud accounts as you need — multiple AWS accounts, GCP projects, Azure subscriptions, and DigitalOcean accounts, all at the same time. Everything appears in a single unified dashboard. Filter views by provider, account, or region.

CloudAIPilot uses read/write permissions scoped to the specific services you want to manage. It uses IAM roles on AWS, Service Accounts on GCP, App Registrations on Azure, and API keys on DigitalOcean. The connection wizard shows you the exact permissions required — you copy them into your cloud console, grant access, and connect. CloudAIPilot never asks for permissions beyond what it needs.

Yes. CloudAIPilot never stores plaintext credentials. All connection tokens and keys are encrypted and stored in a dedicated secrets vault. The AI Pilot uses the principle of least privilege for every operation — it only requests the access it needs for the specific action you approved. You can revoke access at any time by disconnecting the account.

Yes. Disconnecting a cloud account removes the integration immediately and deletes the stored credentials. Your cloud resources — servers, databases, storage — are completely untouched. Disconnecting only removes CloudAIPilot's ability to manage them.

The AI Pilot is a conversational AI assistant built into your dashboard. You describe what you want in plain English — "resize my production server", "why is disk usage so high?", "set up daily backups for this database" — and the AI Pilot generates a governed proposal. You review exactly what will happen, then approve or reject. Nothing executes without your explicit confirmation. Every action is logged in the Activity Center with a full trail.

No. The AI Pilot never applies changes automatically. Every action goes through a proposal dialog — the AI describes what it intends to do, takes a snapshot before making changes, and waits for your explicit approval. You can also reject a proposal or ask the AI to try a different approach. This is a deliberate design decision: the AI Pilot is a governed assistant, not an autonomous agent.

The AI Pilot can be powered by models from Anthropic, OpenAI, or Google. You choose which AI company to use in your organization's AI Pilot settings. Your admin can also lock this setting so all team members use the same model.

Yes. In your AI Pilot settings, you can select the AI company and model you prefer. Organization admins can lock the model selection for the whole team — useful when you want consistency across team members or for cost control. The model setting is audited and logged.

Nothing is executed. The proposal is discarded and logged in the Activity Center as "denied" — with the actor, the proposed action, and the timestamp. You can then ask the AI Pilot to suggest a different approach, ask it to explain its reasoning, or handle the operation manually through the dashboard.

The AI Pilot operates under your RBAC permissions. If a team member's role does not include file or database access, the AI Pilot cannot access those resources on their behalf either. Sensitive operations like database queries, file writes, and shell commands are separately controlled by the agent:execute permission, which you can grant or restrict per team member.

Any instance type available on your connected cloud provider — CloudAIPilot passes your selection directly to the provider's API with no restrictions. On AWS that means the full EC2 catalogue, on GCP the full Compute Engine range, on Azure all VM sizes, and on DigitalOcean all Droplet plans. The provisioning wizard shows available sizes in real-time based on your selected region.

Yes. CloudAIPilot provides a browser-based SSH terminal in the server detail panel — no setup required. You can also use your own SSH client with the key pair you configured during provisioning. For managed servers, CloudAIPilot installs a lightweight agent that enables additional dashboard features like process monitoring and package management.

Three ways: (1) Use the package installer in the server detail view to search and install packages without touching a terminal; (2) Use the browser SSH terminal to run any command directly; or (3) Ask the AI Pilot — e.g. "install and configure nginx on this server" — and it will generate a proposal with the exact commands it will run.

Yes. Server resize is available from the server detail panel. CloudAIPilot initiates the resize via the provider's API — most providers require a brief restart. The AI Pilot can also recommend resizes based on your monitoring data if you're consistently over- or under-provisioned.

Your servers and all cloud resources are completely unaffected. They are yours — they live in your cloud account, not ours. Cancelling CloudAIPilot only removes your access to the management dashboard. Your servers continue running exactly as they were. You will be billed by your cloud provider directly as normal.

CloudAIPilot has first-class support for WordPress, Node.js, Python (Flask, Django, or generic), PHP (generic), and static sites. Any containerized application can also be deployed via Docker or Docker Compose. If your stack isn't in this list, you can almost certainly run it in a container.

CloudAIPilot provisions a Let's Encrypt certificate automatically when you attach a custom domain to a site or app. Once your DNS is pointing to the server and propagation completes, the certificate is issued and configured without any manual steps. Renewal is handled automatically before expiry — you never need to think about it again.

Yes. Connect your GitHub, GitLab, or Bitbucket account and link a repository to your site or app. Deployments trigger automatically on push to the configured branch. You can set build commands, environment variables, and deploy hooks. A visual CI/CD pipeline builder is coming in a future update.

In the site or app settings, add your domain and CloudAIPilot will show you the IP address or DNS record to set in your domain registrar. Once you point the domain's DNS to that address and propagation completes (usually 5–30 minutes), SSL provisions automatically and the domain goes live.

Yes. Upload a docker-compose.yml directly or link a repository that contains one. CloudAIPilot deploys the entire stack on your chosen server, manages container lifecycle, streams logs per service, and lets you restart or stop individual services from the dashboard.

You control the schedule. Options include hourly, daily, weekly, or any custom cron expression. You can configure separate schedules for filesystem snapshots and database dumps on the same server.

Backups are stored in your own cloud storage bucket — S3, Google Cloud Storage, Azure Blob Storage, or DigitalOcean Spaces. You choose the destination, retention policy, and who has access. CloudAIPilot does not store copies of your backups on our own infrastructure.

Yes. CloudAIPilot supports scheduled dumps for MySQL, PostgreSQL, and MariaDB. Dumps are compressed, encrypted, and uploaded to your configured storage bucket. You can restore a specific dump independently of a full filesystem restore.

From the Backups panel, browse your backup history, select the snapshot or dump you want to restore, and click Restore. The AI Pilot handles the restore operation as a governed proposal — it shows you what will be overwritten, takes a fresh snapshot first, and waits for your confirmation before proceeding.

CloudAIPilot's Cloud Storage module supports Amazon S3, Google Cloud Storage (GCS), Azure Blob Storage, and DigitalOcean Spaces — all manageable from the same unified file manager. You can browse folders, upload or download files, manage permissions, and generate public URLs across all four providers from one view.

CloudAIPilot monitors CPU usage, memory, disk I/O and usage, network traffic, uptime, and application-level health checks. Metrics are displayed in real-time charts with historical data. The AI Pilot can also analyze trends — for example, projecting when disk usage will reach a critical threshold based on current growth rate.

Alert notifications can be sent to: Slack, email, Discord, PagerDuty, Microsoft Teams, webhooks (any endpoint), and SMS (extra charge applies). You can configure multiple channels per alert rule and set different channels for different severity levels.

Yes. Each alert rule lets you configure the metric, threshold value, duration (how long the threshold must be exceeded before alerting), severity level, and escalation rules. You can set multiple rules for the same metric — for example, a warning at 80% CPU and a critical alert at 95% CPU, each routing to different channels.

Yes. The AI Pilot scans monitoring data for anomalies and surfaces insights proactively — things like unusual traffic spikes, memory creep over time, or disk usage approaching capacity. You can also ask it directly: "what's going on with this server?" and it will analyse the recent monitoring data and give you a plain-English summary.

CloudAIPilot includes five built-in system roles: Owner (full access — all 46 permissions), Admin (44 permissions — most operations except organization deletion), Member (20 permissions — day-to-day operations without administrative access), Viewer (13 permissions — read-only across most resources), and Agency Partner (a custom role designed for agencies managing client organizations with scoped access).

Yes. You can create custom roles from scratch or start from one of five built-in role templates and adjust individual permissions. CloudAIPilot has 46 permissions across 14 categories — covering infrastructure, sites & deployments, backups, storage, monitoring, teams, billing, AI Pilot, and more. Each permission can be toggled on or off independently.

IP allowlisting lets you restrict a team member's dashboard access to specific IP addresses or CIDR ranges. If that team member tries to access CloudAIPilot from an IP address not on their allowlist, access is denied — even if they have valid credentials and MFA. This is configured per member, not per role. Organization owners are never blocked by IP restrictions.

Yes. Set an access expiry date on any team member's account from the Members panel. When the expiry date and time passes, their access is automatically revoked and they cannot log in — no manual action needed. This is particularly useful for contractors, freelancers, or temporary collaborators. Expired access is logged in the Activity Center.

FinOps (Financial Operations) is the practice of understanding, controlling, and optimizing cloud spend. CloudAIPilot's FinOps module pulls cost data from your connected cloud providers, normalizes it into one dashboard, and uses AI to surface waste, anomalies, and savings recommendations — so you don't need to log into four different provider consoles to understand your cloud bill.

Via each provider's native billing API — AWS Cost Explorer, GCP BigQuery Billing Export, Azure Cost Management API, and DigitalOcean Billing API. CloudAIPilot reads cost data only — it never modifies your billing configuration or payment methods. The same connection credentials used for infrastructure management cover billing access.

Cost data is pulled once daily at 04:00 UTC. Most cloud providers have a 24–48 hour lag in their own billing data, so figures may trail slightly behind real-time usage — this is a provider-side limitation, not CloudAIPilot's. AI recommendations and anomaly detection run automatically after each cost pull.

CloudAIPilot supports USD, EUR, GBP, INR, JPY, AUD, and CAD as display currencies. Exchange rates are sourced daily from the European Central Bank (ECB), with exchangerate.host as a fallback. All cost data is stored in the provider's native currency and converted at display time.

No. The cost agent surfaces recommendations but never applies them on its own. Accepting a recommendation opens the AI Pilot proposal dialog — you review exactly what will happen, the AI takes a snapshot, and waits for your explicit approval. This applies to all cost actions: server rightsizing, resource deletion, reservation purchases, and anything else.

The Savings widget is a running total of the monthly savings achieved from approved and executed cost recommendations. Every time you approve a recommendation through the AI Pilot and it is executed, the estimated saving is added to the widget. It shows the real-world cumulative impact of the cost actions you have taken — not a projected estimate, but actual implemented savings.

There are five tiers. Free is permanently free — 1 server, 1 cloud account, no credit card. Solo starts at ₹749/month ($9/month) and covers 5 servers and 3 cloud accounts. Studio starts at ₹1,999/month ($24/month) for 10 servers and 5 cloud accounts. Agency starts at ₹5,999/month ($69/month) for 75 servers and unlimited cloud accounts. Enterprise is custom pricing for large-scale deployments — contact us to discuss. All paid plans include a 14-day free trial and save around 20% on annual billing.

CloudAIPilot uses flat subscription billing — not per-server or per-user metering. You pay a fixed price for your plan tier each billing cycle. Choose monthly or annual billing; annual saves around 20%. AI Pilot conversations draw from your account's AI credit balance — paid plans include a monthly credit allocation, and you can top up with pay-as-you-go credits if needed.

If you're not satisfied within the first 14 days of a paid plan, reach out to support@cloudaipilot.com and we'll make it right. After the trial window, subscription fees are non-refundable for the current billing cycle. You can cancel at any time to stop future charges — your account remains active until the end of the paid period.

Yes. You can change your plan at any time from your billing dashboard. Upgrades take effect immediately and are prorated for the remainder of your billing cycle. Downgrades take effect at the next renewal date, so you keep your current features until the cycle ends. Cancellations work the same way — your account stays active until the period you've paid for expires.

Yes — each plan has limits for servers, cloud accounts, team members, sites, backup retention, and other features. You can review the full breakdown in the plan comparison table. If you're approaching a limit, you'll get an in-app notification. Upgrading your plan instantly increases your limits. Enterprise plans come with custom limits tailored to your scale.

All data in transit is encrypted using TLS 1.3. All data at rest is encrypted using AES-256. Credentials, API keys, and secrets are stored in a dedicated encrypted secrets vault — never in the application database in plaintext. Encryption is applied to all data — not just selected sensitive fields.

Credentials are stored in an encrypted secrets vault, not in plaintext anywhere in our systems. CloudAIPilot never logs, exposes, or transmits credentials in API responses or support communications. You can revoke access at any time by disconnecting the cloud account — which immediately deletes the stored credentials from the vault.

Post-quantum cryptography (PQC) refers to cryptographic algorithms designed to remain secure against attacks from quantum computers, which could eventually break conventional encryption methods like RSA and elliptic curve. CloudAIPilot implements PQC algorithms alongside traditional encryption for key operations — providing an additional layer of future-proof protection that most cloud management platforms don't offer. This is a deliberate architectural decision, not an afterthought.

CloudAIPilot supports multi-factor authentication (MFA), configurable session timeouts, IP allowlisting per team member (so access can be locked to specific IP ranges), and automatic access expiry for team members. Every login attempt, failed login, session start, and permission change is logged in the Activity Center with full detail — giving you a complete audit trail of who accessed your organization and when.

Yes. Contact support to request a full data deletion. CloudAIPilot will delete your account, organization data, audit logs, and stored credentials within 30 days of the request. Your cloud infrastructure is completely unaffected — only the CloudAIPilot management platform data is removed. A deletion confirmation will be provided once complete.

You can reach CloudAIPilot support via the in-app support chat (in the dashboard) or by emailing support@cloudaipilot.com. The in-app chat includes context about your account and recent activity, which helps us resolve issues faster. Full documentation is available at docs.cloudaipilot.com.

Yes. You can email us at support@cloudaipilot.com for any support queries, billing questions, or general help. We typically respond within a few hours during business days. For fastest resolution, the in-app support chat is recommended as it automatically includes your account context.

Yes — full documentation is available at docs.cloudaipilot.com. It covers getting started guides, cloud account connection setup, feature-by-feature walkthroughs, the AI Pilot reference, RBAC permission definitions, and the API reference for developers building on top of CloudAIPilot.