Our Privacy Philosophy
Your data works for you.
Not for us. Not for anyone else.
We build AI that remembers you, learns your business, and gets better over time. That only works if you trust us with what you share. Here's how we earn that.
Plain language first. Legalese later.
Before the formal policy below, here's what we actually believe and how we actually operate — in words a human would use.
Your data is yours.
Everything you put into a Kin system — conversations, documents, knowledge bases, business context — belongs to you. We store it to make your system work. That’s it.
We don’t train on your data.
Your information is never used to improve other customers’ systems, build general models, or feed anything outside your own instance. Your Spark learns about your business. Nobody else’s Spark benefits from that. Ever.
You can take it with you.
Want a copy of everything we store for you? Ask. We’ll export it. Want to leave? Ask. We’ll delete it. No retention games, no 90-day windows, no “we’ll get back to you.” You say delete, we delete.
You can see what we see.
Your data isn’t locked behind dashboards we control. You have access to your knowledge base, your conversation history, your system configuration. If it’s yours, you can reach it.
What we control — and what we don't.
Kin systems are built on large language models from providers like OpenAI, Anthropic, and Google. When your Spark processes a conversation, that conversation passes through their infrastructure.
Here's what that means:
What we control:
- What data we store and how we store it — encrypted, isolated, yours
- What we do with your data — nothing, besides powering your system
- How long we keep it — as long as you want, not a day longer
- Who sees it on our end — only what's necessary to keep your system running
- Our architecture choices — we select providers and configurations that minimize data exposure
What we can't control:
- How LLM providers handle prompts that pass through their APIs
- Whether those providers change their policies in the future
What we do about it:
- We use API configurations that opt out of training wherever providers offer that option
- We stay current on provider privacy policies so you don't have to
- We architect systems to send the minimum necessary context — not your entire knowledge base on every request
- When better options emerge — local models, on-premise deployment, privacy-first providers — we move toward them
We won't pretend we've solved a problem that lives upstream from us. But we will tell you exactly where the boundaries are and what we're doing to push them in your favor.
Privacy isn't a feature. It's the foundation.
The whole point of what we build is AI that knows you. That understands your business from the inside. That remembers.
That kind of depth requires trust. And trust isn't a terms-of-service checkbox — it's a relationship. You share something real with us when you let us into your business voice, your customer interactions, your internal knowledge.
We take that seriously. Not because regulations say we have to — though they do, and we comply — but because the moment you stop trusting the system is the moment it stops being useful.
So we start from a simple place: your data exists to serve you. Everything else follows from that.
Your rights, no asterisks.
Access
Request a full export of everything we store for you, anytime.
Deletion
Request complete deletion of your data, anytime, no questions asked.
Transparency
Ask us exactly what data we hold and how it’s being used.
Portability
Your knowledge base, your conversation logs, your configurations — exportable in standard formats.
Correction
Something wrong in your data? Tell us, we fix it.
No Surprises
If anything about how we handle data changes, you’ll hear about it from us directly, in plain language, before it takes effect.
Get in touch.
Questions about your data? Something feel off? Just want to know more?
We don't route privacy questions through a support queue. You'll hear from a person.
Formal privacy policy coming soon.
In the meantime, everything above is how we actually operate. Questions? Reach us at kin@kinetiq.marketing