Questions we get asked

A secure computing environment where approved researchers can analyse sensitive data without being able to copy it out. Data comes in through a controlled route, analysis happens inside, and only checked and approved results leave. It follows the Five Safes: safe people, safe projects, safe settings, safe data and safe outputs.

Small and mid-sized research organisations that handle sensitive data, such as health, genomic-adjacent or personal data, and don't have a DevOps team to build and run their own TRE.

No. TrustBoxx is a managed service. Techvito builds, monitors, patches and backs up the environment. Your team only needs to do research and decide who gets access and what can leave.

A new environment is deployed in about two days once we have agreed the setup and your identity details. Loading your data and onboarding researchers usually takes a few days more.

You can, and open-source options exist. In practice it takes a capable infrastructure team weeks to months, and then someone has to run it. See build vs buy for a full comparison.

Not yet. Certification is on our roadmap and Cyber Essentials is in progress. What we offer instead is an architecture you can inspect, a published SATRE self-assessment, a full audit trail, and a live production deployment handling regulated UK healthcare data.

We have self-assessed against SATRE 2.0 and publish the results, including the statements we do not yet meet. See our SATRE page.

No. Every client gets a dedicated, single-tenant AWS environment. Nothing is shared with other clients.

Only in the AWS region you choose, for example London for UK data or Mumbai for Indian data.

Clipboard, downloads and local drive mapping are blocked, and all activity is logged. As with any remote desktop, a determined user can photograph their screen. That is why TrustBoxx combines technical controls with researcher accreditation, monitoring and the outputs workflow.

Yes. Backups are part of the managed service and stay in the same region as your environment.

Researchers submit files from inside the TRE. Automated disclosure checks run first, then a named approver outside the TRE approves or rejects. Approved files are released through a short-lived link and every decision is logged. See safe outputs.

Through a controlled ingress route. Each file is hashed and its acceptance recorded before it reaches the workspace.

We recommend that data is pseudonymised by the data owner before it enters TrustBoxx. We can advise on your approach.

You do. Your code, results and intellectual property belong to your organisation.

Approved outputs are released through the normal workflow, then the environment is destroyed, including storage and backups. We confirm destruction in writing.

R, RStudio, Python, Jupyter and VS Code, with Gitea for version control and an internal repository of approved packages.

Yes. You provide the licence and we install and configure the software inside the environment after review.

Yes. Private AI gives researchers an LLM assistant inside the environment, through AWS Bedrock. Data and prompts stay in your environment and are not used to train models. See Private AI.

GPU workspaces are on our roadmap and not generally available yet. If you need them for a project, talk to us about timing.

Each environment belongs to one client organisation. Talk to us about how to give named access to collaborators.

A one-time setup fee, a fixed monthly platform fee per environment, and cloud usage passed through transparently. Every project gets a written cost estimate before work starts. Pricing is on request.

TrustBoxx is built and operated by Techvito, a software products company.

Our first production deployment is with a UK healthcare research organisation. We can't name them publicly, but they can provide a private reference on request.