# Safe cloud deployment for AI-built software Salus helps IT leaders deploy team-built apps and agents, control access, and monitor them in one place. [Try now ](https://app.salus.cloud/register) No card needed · [up to $25 in credit as you set up](https://salus.cloud/pricing/#credits) SOC 2 Type II and ISO 27001 compliant Illustration: an isometric city at the blue hour. Across a canal from the old town, where core systems keep their own light, dark, unmanaged buildings stand under a label, "Shadow AI": apps and agents teams built with AI, outside IT's view. One is wedged against the core-systems power station, others sit on rented lots floating off the ground and on scaffolding at the edge. Then IT's watchtower lights and its beam sweeps slowly across the scene. Behind it a survey grid draws the district; each shadow building lights up where it stands and lifts, a sign on the old town's bank comes up reading "Core systems", then one after another they glide into plots that carry the organization's rules, and land. New apps then arrive over the bridge, by hand or by a builder's drone, through the same inspection gate. The district holds five teams' spaces: Finance, HR, Ops, Marketing and Customer success. At its edge a sign reads "Salus GCP · USA": the whole city is one Salus Cloud region. Its road crosses the river, a faint lime line of the same rules running along it, to a signpost pointing to Salus AWS (South Africa) and to your own cloud on Enterprise. Illustrative product view AcmeProductionSalus GCP (USA) Organization rules Applies to every space · 5 spaces · 12 workloads - Scan before release - Roles - Audit log Finance0 critical - Budget dashboardApp - Invoice checkerAgent - Spend approvalsApp HR0 critical - Onboarding portalApp - Policy assistantAgent Ops0 critical - Inventory trackerApp - Shift plannerApp - Ticket triageAgent Marketing0 critical - Campaign siteApp - Lead scorerAgent Customer success0 critical - Renewals trackerApp - Support insightsAgent Activity - **Priya N.** deployed Budget dashboard · 2m - **Sam K.** deployed Ticket triage · 9m - **Lerato M.** deployed Campaign site · 24m - **Tom W.** took Shift planner offline · 1h - Data, workloads and resources stay in Salus GCP (USA) - Separate from core systems - p95 latency 38 ms · in region - 0 critical findings - ![Yoco logo](https://salus.cloud/assets/logos/yoco.svg) - ![Mukuru logo](https://salus.cloud/assets/logos/mukuru.svg) - ![Deimos logo](https://salus.cloud/assets/logos/deimos.svg) - ![Yuppiechef logo](https://salus.cloud/assets/logos/yuppiechef.svg) - ![PiggyVest logo](https://salus.cloud/assets/logos/piggyvest.svg) - ![Zedcrest logo](https://salus.cloud/assets/logos/zedcrest.svg) ## Everyone is building apps Finance, HR, ops, and marketing are constantly shipping apps and agents with AI. You need a way to manage what they put into use. ### Teams build tools Employees use AI to create tools for their teams, from budget dashboards to support apps to full AI agents. ### New tools need hosting Hosting new software beside core systems adds risk. Separate team accounts make oversight harder. ### IT inherits the rest IT leaders are tasked with handling setup, security checks, and maintenance alongside any existing work. ## Give teams an approved way to deploy 1. ### 1\. Set your rules Set permissions across your organization, spaces, and projects, and release rules new projects inherit. 2. ### 2\. Deploy the app or agent Teams connect the apps and agents they build. Salus deploys them through your rules. OpenAI Codex · Salus MCP Works with your favorite agent Illustrative 1. Priya N. · Finance: Deploy the expense tracker. 2. `deploy_repo` expense-tracker 3. Using Priya N.’s access 4. `get_pipeline` Scan ▸ Build ▸ Release 5. Running in Salus GCP (USA) 3. ### 3\. Review and monitor anything Check logs, health, and usage from one place. Take a workload offline when needed. Cursor · Salus MCP Works with your favorite agent Illustrative 1. Tom W. · IT: Anything failing in production? 2. `get_deployment_overview` production 3. `get_app_logs` shift-planner 4. `get_deployment_metrics` shift-planner 5. Shift planner: error rate up since 14:02 6. Take Shift planner offline. 7. `set_deployment_offline` shift-planner: Shift planner is offline 8. Bring Shift planner back online. 9. `set_deployment_online` shift-planner: Shift planner is running Illustration: the same city as at the top of the page, seen from higher up. First, IT's rules land: a lime boundary is drawn round the teams' district and a lime code mark drops onto every plot, including an empty one, because new projects inherit the organization's rules. Then a builder's drone carries a new app, Expense tracker, over the bridge to the inspection gate, where it is scanned, then built on its plot and released, its windows lighting up. Last, IT's watchtower turns its beam across the district; one workload, Shift planner, is taken offline and goes dark, then is brought back online. Beside each step, an illustrative agent session shows the same moment asked of an agent through the Salus MCP server. ## Run team apps with less upkeep Safer deployment ### Reduce risk to core systems Give team apps and agents separate hosting so they run apart from core business systems. Separate managed hosting and vulnerability scanning Control ### Control who can deploy Apply shared rules as teams add software. Review access and activity through the same platform. Team permissions and activity records Capacity ### Reduce hosting work Use managed infrastructure for new apps so your team has less hosting to set up and maintain. Cloud infrastructure and app monitoring Spending ### Track usage as teams grow View usage by app so you can see what consumes your prepaid balance. Resource metering by app ## Control deployment and daily use Deploy team apps and agents with shared controls, managed hosting, and monitoring in one platform. 01Automated deployment02Agentic control03Governance rules04Vulnerability scanning05RBAC06Cloud infrastructure07Logging08Metrics09Metering ### Automated deployment Connect the code your teams build and get it running through your release rules. Use the same process when they update an app or deploy a new agent. Illustrative product view Expense tracker ScanBuildRelease Deployment running ### Agentic control Connect a builder’s own agent through MCP to deploy software, read logs, or roll back a release. It uses the platform permissions of the person who invokes it. Illustrative product view Your builder’s agent Salus via MCP Request a deployment Uses the builder’s platform access Deployment • Logs • Rollback ### Governance rules Set release rules once for your organization. New projects inherit those controls, so your team spends less time setting up each project. Illustrative product view Organization release rules Budget dashboard Shift planner New projects inherit the rules ### Vulnerability scanning Check software for known vulnerabilities before release. Review the findings as part of deployment so your team can address risks before putting an app into use. Illustrative product view Vulnerability report | Package | Severity | Finding | | --- | --- | --- | | Dependency A | Review | | Dependency B | Review | | Dependency C | Review | ### RBAC Choose who can view, change, and deploy workloads across spaces and projects. Give each team the access it needs while keeping platform permissions under IT control. Illustrative product view Project permissions | Role | View | Change | Deploy | | --- | --- | --- | --- | | IT admin | ✓ | ✓ | ✓ | | Team builder | ✓ | ✓ | ✓ | | Viewer | ✓ | — | — | ### Cloud infrastructure Run team apps and agents on managed hosting separate from core business systems. Reduce the infrastructure your team needs to set up and maintain for new software. Illustrative product view Core business systems Separate managed hosting - Team apps and agents - Databases Separate environments ### Logging Read workload logs through the platform when an app or agent needs attention. Investigate what happened using the records from the software you already run. Illustrative product view Expense tracker / Logs TimeServiceMessage 12:04webRequest received 12:04apiTask started 12:05workerTask completed 12:06apiExample error ### Metrics Check workload health and performance through the platform. See how your apps and agents respond during use so your team can investigate changes in behavior. Illustrative product view Requests Errors Response time Workload performance over time ### Per app metering See which apps drive platform usage on Salus Cloud. Track what consumes your prepaid balance so you can make spending decisions with usage in view. Illustrative product view Salus Cloud usage Prepaid balance By app By resource Compute · Memory · Storage · Bandwidth Select a feature to see its product view ## Support any builder or leader I run ITI build team toolsI run infrastructureI manage the budget ### I run IT Give teams an approved place to deploy apps and agents. Set shared rules and review their workloads from one platform. [Try now ](https://app.salus.cloud/register) ### I build team tools Deploy what you build within the rules IT sets. Share your apps with colleagues without extra app user fees. [Start free ](https://app.salus.cloud/register) ### I run infrastructure Keep new team workloads separate from core systems. Use managed hosting with logs, metrics, and vulnerability scanning. ### I manage the budget See usage by app, project and environment. Support new software through a managed platform instead of funding an internal hosting setup. Cloud ## Choose your cloud ### Salus Cloud Start on managed Salus Cloud, and choose the cloud and region each environment’s workloads and data run in. ### Your own cloud Go Enterprise to bring your own cloud for AWS, GCP, or on premise. ### Cloud portability Move full stack apps between supported environments as your hosting needs change. Ask your agent to make the move. [Learn more about Enterprise](https://salus.cloud/enterprise/) Illustration: the horizon of a small dark planet, a thin lime atmosphere along its curve for IT's rules. On it stand three districts: one of your own cloud accounts, with blue roofs, and the two Salus Cloud regions, Salus GCP (USA) and Salus AWS (South Africa), each with lit windows. Faint lime tethers run from the Salus control plane to the other districts. Pricing ## Compare plans ### Hobby $0 plus usage Get started on Salus Cloud with usage charges for running workloads. [Start free](https://app.salus.cloud/register) ### Pro $9 per builder / month, plus usage Run software on Salus Cloud with paid builder seats plus usage charges. [View pricing](https://salus.cloud/pricing/) ### Enterprise From $25 per builder / month, plus usage Stay on managed Salus Cloud or add bring your own cloud for your company requirements. [Book a demo ](https://salus.cloud/talk-to-sales/) FAQ ## Questions IT leaders ask ### What is Salus? Salus is a safe cloud deployment platform for AI-built software. IT leaders use it to give teams an approved way to deploy the apps and agents they build, control who can deploy, and monitor them in one place. A builder connects a GitHub, GitLab or Azure DevOps repository, and Salus scans its packages for known vulnerabilities, builds the app as it is and deploys it as a container under the release rules IT sets. Plans start at $0. ### How much does Salus cost? Salus is priced per builder seat, plus usage for what runs. Hobby is $0, Pro is $9 per builder per month, and Enterprise starts at $25 per builder per month. A seat is a builder or an admin. The colleagues and customers who use what your teams deploy are never billed and don’t need a Salus account. Running workloads in your own cloud (BYOC) is an Enterprise option, priced per cluster. ### Where do apps deployed with Salus run? On managed Salus Cloud, you choose the cloud and region for each environment when you create it. Today that is Google Cloud in the USA or AWS in South Africa. On Enterprise, you can also deploy into a dedicated, isolated environment in infrastructure you own: an AWS, Google Cloud or Huawei Cloud account, a private cloud, or servers in your own data center. Your agent can move a full-stack app to another supported environment you have already set up. ### How does IT stay in control of what teams deploy? IT sets the rules once and deployments follow them. Roles apply at three levels, organization, space and project, so a viewer can’t deploy. Release rules set for the organization are inherited by every new project, and every pipeline scans for known vulnerabilities before release. Deploys appear in the audit log under the builder whose access was used, and IT can take any deployment offline. ### Can AI agents deploy with Salus? Yes. A builder can connect an AI coding agent such as Claude Code or Cursor to Salus through its MCP server, which has 52 tools for deploying, reading logs, managing databases and rolling back. The agent acts for the builder who connected it, with that person’s access, and its deploys go through the same release rules and vulnerability scans as anyone else’s. Agents your teams build deploy and run like any other app. ### Is Salus SOC 2 and ISO 27001 compliant? Yes. Salus is SOC 2 Type II compliant and ISO 27001:2022 certified, both following independent audits, and the SOC 2 Type II report is available under NDA for your security review. Independent third-party penetration testing is in progress. Team apps run on managed hosting separate from your core systems, sign-in uses short-lived tokens, and audit log entries are stored append-only and isolated per tenant. ## Ready to give your teams an approved way to deploy? See how Salus handles deployment, access, and monitoring for the apps and agents your teams create. [Try now ](https://app.salus.cloud/register) No card needed · [up to $25 in credit as you set up](https://salus.cloud/pricing/#credits)