ContentGrid

ContentGrid calls itself a Content Service Platform. It is aimed at IT and operations teams in insurance, banking and HR that need document-heavy applications such as archives and partner portals, more than at people building a marketing website. The service is run by Xenit, which is how the company's own pages refer to the provider, and the pages do not state a headquarters location.

What you build and how

The site describes the workflow in four steps. You model your content "using a business information model" with "No folders needed." You then build tailored content applications with AI features such as auto-classification and data extraction, deploy them "across preferred cloud platforms," and secure and scale access and storage to match your needs. The homepage promises to turn "a forest of files into a structured landscape," free from folders and "role explosion" thanks to a rich relational model and transparent policy-based access control.

The model is the center of the product. Instead of dragging blocks onto a canvas, you define entities and the relations between them, and the console generates the content application from that model. The site's feature names say the same thing: Meta Model Driven, Scales, AI First and Protects By Policy. The generated application is described as simple code that relies largely on the Spring Framework and Spring Boot, with Postgres as the database. Because the generated code is plain Java on open frameworks, developers can read it, extend it and call "the full and documented (and protected) API."

The use cases the homepage lists give a clear picture of what teams build:

1. Automated inbound document classification.
2. Content portals for partners, suppliers, employees and customers.
3. Legal archiving.
4. AI-based document analysis.
5. Automation of content-driven business processes.

The site groups its solutions under Use Cases and Industries, with a dedicated page for insurance, and a blog post argues that traditional enterprise content management and role-based access control will not carry insurers through DORA and eIDAS v2. The proposed answer is policy-based and attribute-based access control, so that sensitive documents stay secured and data handling stays traceable and auditable across systems. If you have tried to model who may see which contract with folder permissions, that argument will sound familiar.

AI shows up in three places. The first is the classification and extraction step in the build flow. The second is an AI Assistant: an August 2026 recap post says the platform now has "a real AI Assistant and full-text search now built in." The third is the positioning as an "AI-friendly ECM system" with smart categorization and predictive insights. The recap also mentions access management that the customer's team controls without waiting on the vendor.

On scale, the company published a performance test on a dataset of 500 million documents, looking at query response times and throughput under varying concurrency. Another post introduces content encryption for secure document storage that still supports range requests. The pages do not give the test results in numbers, so read the post itself if scale is your deciding factor.

For connecting other systems, the homepage talks about rapid integration and a scalable content platform with a user-friendly interface. A partner page lists a Salesforce x ContentGrid combination, and the navigation includes a partner program and a page on Woo (Wet open overheid). The pages name no other connectors, and no template count.

The evidence from the pages

Most of the hard facts sit on the no-vendor-lock-in page, and they are unusually specific.

• License. "All software required to run your app is released in the most liberal open source license, namely apache 2.0." The generated content applications are open source, and the customer is free to share the sources with any party. The company also links an open source project called thunx on GitHub.
• Console escrow. The console software "is placed under escrow," so the code is available if the company running the service were to default.
• Hosting. ContentGrid is "a SaaS first offering" but supports "bring your own cloud" architectures. SaaS customers who want to move their run-time app to a public cloud such as AWS or Azure are supported to do so. With bring-your-own-cloud you can disconnect from the central console at any time without impact on production. The monitoring platform, Grafana Cloud, still needs an open connection to deliver reporting.
• No license keys. "We don't do license keys. We do service." If you stop your subscription and payments, the SaaS service ends after 3 months, or as long as it takes to extract your content and metadata, though a self-run service keeps running without updates.
• Exit. Three processes start on exit: an export of content and metadata through the API, a copy of the console so the app can be generated and maintained, and a move of the service to the customer's own cloud. The exit plan is "tested" on a yearly basis.
• Training. The page describes "free Grid days," run once a year, where major customers can have engineers or architects trained on the platform.
• Partners. A partner network of co-creators and service providers is described as able to run, service and adapt apps.

On money, the site publishes no prices in the pages held here. It does state that it dislikes "opaque pricing models" and that the evolution of pricing is meant to be predictable and open. The navigation links a Pricing page and a Free Demo, so expect to confirm the numbers there. The page does not mention a free plan.

On customers, a CTO at RGF NL is quoted, and the customer stories also name Matthias of Patronale Live. One testimonial on the page says "A lot of costs are saved by working with an efficient document management system," though the page does not make clear which of the two said it.

Who it suits and who it does not

ContentGrid suits an IT or enterprise team in a regulated sector that needs a governed document platform it can eventually own. The Apache license, the escrow arrangement, the exit process and the bring-your-own-cloud route are all answers to one concern, which is not being trapped. The policy-based access model suits organizations where "who may see this" depends on attributes of the document and the person, not on a folder.

A developer who wants an escape hatch into code gets one here, since the generated app is open source Java and Spring.

The caveats follow from the same design. This is not a drag-and-drop website builder, and it does not describe themes, page templates or a marketing site editor. Defining a good information model is up-front work, and the open source runtime implies someone on your side or a partner's side who can operate it. A solo founder or a small team that wants a blog live this week will find a different tool a better match. And because pricing is not shown in these pages, budgeting means starting a conversation with the vendor.