Contember Fabrika is a platform that lets companies build their own applications safely with AI, and it consists of the platform itself and the team behind it. The platform handles login, permissions, and deployment, and all changes go through review.
Purpose: Companies can use AI to quickly build internal applications, such as approval workflows, contract registries, field records, invoice approvals, sales pipelines, project timesheets, customer portals, order management, capacity planning, and more. The platform is designed to solve the security, permissions, deployment, and maintenance problems caused by AI-generated code, and to avoid shadow IT.
Main capabilities:
- Unified entry point: Applications do not have independent public addresses; everyone logs in through the platform.
- AI internal building: Builders use AI to create interfaces and logic based on requirement specifications and code, without touching live data.
- Secret isolation: Passwords and keys are kept by the cloud platform, and each application can only see its own keys.
- Audit records: Logins, permission changes, and every deployment are recorded, and the console shows who did what and when.
- Change review: Every deployment is reviewed by the team before it goes live; anything that does not pass does not go live.
- Platform features: Support for logging in with company accounts (such as Google or Microsoft), roles and permissions, application keys and passwords, deployment from Git, and records of logins, permission changes, and deployments.
- Operations support: The team reviews every change, and each version comes with an inspection report; patches, incident monitoring, and help are provided.
- Infrastructure: Cloud accounts are set up in the company's name, and Cloudflare, Zerops, or AWS can be used; code, data, and domains belong to the enterprise, and the system can run independently of the platform.
How to use it: Fabrika has three onboarding paths:
- Build the first application together with the team: Suitable for companies starting from scratch. On Monday, work through one process with the process owner; from Tuesday to Thursday, build; on Friday, the application officially goes live. Future application builders participate throughout and learn.
- Take over an existing application: The application may have been built by Lovable, Cursor, Claude Code, a freelancer, or an agency. The team reviews the code and makes small changes so the platform can take over login, permissions, keys, and deployment, then helps deploy to Cloudflare or Zerops, with data migrated as well. The prerequisite is that the code can run outside the original tool; applications built with click-based tools such as Bubble cannot be taken over.
- Build with your own developers or agency: If you already have in-house or vendor developers, they build on Fabrika, while operations, change review, and SLA are handled by the platform team.
In any case, the final application runs in the enterpriseaccount, and the platform handles login, permissions, and deployment, with every change reviewed.
Applicable scenarios:
- The company already has people building applications with AI, but no one manages access permissions, and the application dies when the author leaves.
- A department needs a tool, and IT needs to approve and endorse it, but there is no development capability.
- Critical processes rely on Excel and email, and versions are chaotic.
- The system can only be modified by one agency, and changes are slow.
Limitations:
- Applications built in click-based tools such as Bubble cannot be taken over.
- No price list is published; pricing is determined in a quote after scope discussions and then remains unchanged; there is no per-user or hourly charging.
- Kickstart capacity is 2–3 per month.
- Platform setup requires some work, including deploying the platform in the cloud account, establishing AI usage rules, and brief training.
- Specific details such as pricing, supported languages, and platform availability are not yet provided.
