Enjo

GDPR overview

GDPR compliance is stated explicitly on the homepage, the security page and the demo page. Transfers from the EEA, the UK and Switzerland to the United States rely on the European Commission's Standard Contractual Clauses; AWS regions are offered in the US and the EU, and data residency can be requested from the vendor's privacy contact. Access, rectification, erasure, restriction, portability, objection and consent withdrawal are documented with a 30-day response target, although for end users the client organisation is the controller and handles requests, with a right to complain to a local supervisory authority. A published DPA is incorporated into the terms and covers the SCCs, technical measures and the subprocessor list, with 30 days' notice before any new subprocessor. Two gaps: no Article 27 EU representative is named, and the security page still invokes the EU-US Privacy Shield, invalid since 2020. Privacy policy dated 20/03/2026.

Who owns the data?

Ownership is unambiguous. Section 5.1 states that the customer retains all right, title and interest in its Customer Data and that nothing in the agreement transfers ownership to Troopr Labs. Section 5.2 grants the vendor a limited, non-exclusive licence usable only to deliver and maintain the service and to meet legal obligations, expiring when the data is deleted. Section 6.5 assigns AI Outputs generated from the customer's data and prompts to the customer, subject to underlying third-party rights and the AI providers' terms. The privacy policy adds the regulatory framing: for the SaaS products the publisher acts as processor and the client organisation as controller, and is controller only for visitors to its own websites.

Reuse rights

Customers can reuse their own data and the AI outputs it produces without asking Troopr Labs for permission: ownership never moves, subject only to third-party rights and the underlying AI providers' terms. On the vendor side, use is contractually narrow. The privacy policy states that Customer Data is not used to train, fine-tune or improve any AI or machine learning model, is not sold, is not used for advertising and is not passed to AI vendors for training purposes. Only anonymised, aggregated metadata such as usage counts, error rates and latency feeds service improvement, never message content. Inference runs through API access to OpenAI, Anthropic and Microsoft Azure OpenAI under agreements that prohibit training and retention beyond processing the request, each covered by a DPA. Retrieval works by chunking documents into embeddings, described as one-way transformations, stored on AWS. AI interactions may be logged for quality, debugging and audit, with retention configurable by the customer and logging disabled on request. PII detection and masking can be applied before prompts reach the model, though on a best-efforts basis using pattern matching and named-entity recognition. Named recipients include the LLM providers, AWS, Slack, Microsoft Teams, Google Workspace, Google Analytics and Stripe.

Data retention & training

Hosting summary

Enjo runs primarily on AWS, with US and EU regions available depending on the customer's configuration; the security page also names Azure as execution infrastructure, and Azure OpenAI is the default inference provider. The vector embeddings built for retrieval are stored on the AWS infrastructure. The published subprocessor table lists AWS (US and EU), OpenAI (US), Anthropic (US), Microsoft Azure OpenAI (US or EU depending on configuration), Sentry (US) and Stripe (US). The publisher is incorporated in the United States, so transfers from the EEA, the UK and Switzerland are covered by Standard Contractual Clauses, and data residency can be arranged on request through the vendor's privacy contact. Protection measures are documented: AES-256 encryption at rest, TLS 1.2 or above in transit, keys managed in AWS KMS and rotated annually, encrypted daily off-site backups on Amazon S3, and multi-tenant datastores where each customer is isolated by a unique tenant token. In practice a European buyer can obtain EU residency for storage, while inference and several subprocessors remain US-based unless configured otherwise.