A data security company, with software that practises what it sells.
Divelai sanitizes corporate data before it reaches AI models, vendors and offshore processors, and is in closed beta. Its customers trust it with their most sensitive data, so the systems it uses to manage and support those customers had to meet the same standard as the product.
The brief
Divelai's buyers are security, compliance and data teams, and they read the security page before the pricing page. The company needed a site that explains a technical product to them plainly, and software to run the business behind it: one record of every prospect and customer, and a support channel customers can use without their sensitive data ending up somewhere it shouldn't.
What made it hard
- A support request is where customers paste what they shouldn't: logs, payloads, sample records.
- Buyers who judge a security vendor by how it handles their own data, not by what it claims.
- A technical product, from detection to four ways of sanitizing a value, for readers who check every word.
What we built for them
A CRM and customer management system of its own
We built Divelai a custom CRM and customer management system, separate from its website, that the company runs its current clients on. Every client, and every conversation with them, lives in one system Divelai owns rather than in an off-the-shelf tool holding its buyers' details.
Support that sanitizes what customers send
Customers send logs, records and payloads with their support requests. So the support system we built applies the same sanitization principles as Divelai's own product: sensitive values are detected and sanitized, so the people working a request see what they need to solve it and not the data behind it. A customer's support conversation is protected the way their production data is.
A site that answers a security team's questions
The website explains how Divelai detects sensitive data, the four ways it sanitizes a value, redaction, reversible tokenisation, format-preserving masking and synthetic substitution, and how it deploys, with pages for each use, from AI APIs and analytics pipelines to offshore processing and sharing data with vendors. A live demo on the home page shows a request being sanitized, in the visitor's own browser. Every page is built to WCAG 2.2 AA, with zero automated accessibility violations across the site.
Enquiries handled like customer data from the first message
Every access request, quote, security report and privacy request comes through one form, so personal data arrives in one place and can be deleted from one place. Enquiries are encrypted before they are stored, a visitor's IP address is never kept as-is, only a one-way hash, and the form asks people not to paste real customer data. Divelai's team works the queue from a private console where they can search, filter, export and delete.
Want your business built this carefully?
Tell us what you'd like the site to do. A person replies within one business day.
Get a quote