Strongwall.ai is a privacy-first company - we believe everyone has the right to private conversations. The best way to keep information private is for it to not exist, so we do our best to keep as little information about our customers as possible, and what we do need for as short of a time as possible. Every piece of information we have has a concrete reason for existing, not just a vague "to improve the service" justification. And generally, there is a way for customers to avoid giving us that information if they want to.
Let's start with what we don't store.
Storing conversations is a no-go red line for us. Your conversations are never logged or retained. They do not exist in backups. We forward them to and from our GPU servers in such a way that they are only ever in ephemeral memory and immediately discarded after processing. When a customer uses the "sync conversations" feature, there is a brief (usually a matter of seconds) interval where an encrypted copy of your conversations is forwarded through our servers. We cannot read them, and they only exist in our system for as long as it takes them to be downloaded, or a 5 minute timeout, whichever is shorter. You control when syncs happen, and if you prefer, you can export and import your data locally.
Even on your local computer or smartphone, your conversations are stored encrypted in the browser. That means that even on a shared computer, no one can read your conversation history unless they also have your login information.
We don't need to know your identity. We support using an anonymous account number for more privacy-focused customers, as well as anonymous and pseudonymous payment methods.
So what information do we need to run the site?
We do keep track of payments information (when and how much we have been paid), for obvious reasons. If you elected to use a credit or debit card, our payments provider keeps track of the billing information. We fully support prepaid cards.
We also store usage information (essentially, size of conversations, but not their content) so we can determine how many compute resources we need. We store 24 hours of raw usage, a month of daily usage, and then monthly usage after that. This is a rolling window – if someone had a single chat today, then after 31 days, we would only know something like "this customer used 3000 tokens in December 2025".
We also use Google Analytics to determine which of our marketing partners are most effective. This does not give us information about your identity or activity on other sites. Google provides a tool to disable this for your browser entirely.