GDPR and DPA
We sign a data processing agreement when personal data is involved. Processing is limited to the purpose you define.
The rules apply to every project, not only to regulated industries.
We sign a data processing agreement when personal data is involved. Processing is limited to the purpose you define.
Confidentiality is signed before we see code, credentials, or customer data.
Each engineer gets only the access needed for the task. Access is removed when the work ends.
We do not copy production data to local machines. Test data is synthetic or anonymized.
Input validation, prepared queries, nonce and capability checks, escaped output, dependency review.
Client data is not sent to public AI tools without written approval. RAG systems run on your infrastructure or a private environment.
Standard paperwork for an outsourcing engagement.
Project data lives in the client’s systems wherever possible. When we need a working copy, it stays in an access-controlled environment with named users, and it is removed at the end of the engagement. Backups are encrypted, and credentials are stored in a password manager rather than in documents or chat.
For AI features, we separate the model from the data. Retrieval indexes are built from documents you own, hosted in your environment or in a private cloud account. We do not use client content to train public models. Prompts and answers can be logged for quality control only when the agreement allows it.
We are a Ukrainian individual entrepreneur, not a certified ISO or SOC 2 organization. If your procurement requires a certification we do not hold, we say so before the contract, not after.
Procurement and security questions we receive most.