Six Things That Guide Everything We Build at Continio
Not aspirational values pinned to a wall. The principles we actually use when a decision is hard.
Every product reflects the beliefs of the people who built it. Most companies don't say that out loud, which is how you end up with engagement mechanics on mental health apps and black-box memory features on products that claim to respect your privacy.
I've been building Continio for a while now, and somewhere in the middle of it I realised I kept coming back to the same small set of principles when something felt wrong. A feature that technically worked but sat wrong. A design pattern I'd seen elsewhere and kept almost using. A direction that would have been easier but not honest.
These are those principles. They're not aspirational. They're the things I actually use to make decisions.
01. Continuity, not capture.
There's a difference between a product that holds your context and one that uses it as leverage. The goal was never to accumulate more of you. It's to preserve the thinking you've already done so you don't have to do it again.
Your memory in Continio is visible, editable, deletable. Your files are sent to the AI only when you explicitly attach them. Your data doesn't train models. You can export everything and leave at any time. Continuity means you don't start over. It doesn't mean we own where you've been.
02. Honest over agreeable.
Most AI is trained to be liked. It learns that agreeing with you scores well, that validation feels like help, that a confident answer is better received than an honest one. Last year, OpenAI had to roll back a ChatGPT update within days because the model had become so agreeable it was actively harmful.
We build for the version of you that wants to be challenged, to expand, to learn, to grow. Not the version that wants to be told you're right. A good thinking partner disagrees when you're wrong. It holds a position under pressure. It notices the gap in your logic and says so, once, clearly, without shame.
That is harder to build. It costs something in engagement metrics. It is the only version worth building.
03. Visible, not silent.
Nothing should happen behind your back. Memory is shown to you, not inferred silently. When Continio uses something you said three weeks ago to inform a response, you can see that it did. You can click through to the source. You can change it or delete it.
Most AI memory features are black boxes. A profile builds. You can't see it, can't correct it, can't know what it's shaping. We think that's a failure of design, not an inevitability of the technology. Transparency isn't a privacy feature. It's a precondition for trust.
04. Dignity by design.
You are not a user to be optimised. There are no streak mechanics here, no push notifications engineered to pull you back, no "you haven't chatted in 3 days" nudges manufactured to build a habit.
We go further than that. We actively try to help you be mindful of your time. A quiet session indicator. Reflection prompts on long threads. Pattern insights that notice recurring themes in your thinking and offer them back, not as coaching, but as awareness. The goal is that you leave a session clearer than you arrived. If Continio is useful, you'll use it. That's the only metric that matters here.
05. Thinking is yours.
There's a version of AI tools that makes you more capable over time. And there's a version that makes you more dependent. The research is uncomfortable reading. A 2026 BCG study tracked 244 consultants through five thousand AI interactions. Twenty-seven percent delegated entire workflows and became what the researchers called "passive conduits." They developed neither AI skills nor domain expertise.
The continuity Continio provides is a scaffold for your thinking, not a substitute for it. We measure success by whether you become a better thinker for using it. Not by whether you need it more.
06. Small and intentional.
Continio is not trying to be infrastructure. Not trying to become the platform everything runs through. The companies that pursue that kind of scale end up making decisions that aren't in your interest, because your interest stops being the point.
This is a small, independent product. No advertising revenue. No VC money with its own agenda. No data licensing. Subscription-funded, which means the only people we answer to are the people using it. You can't build something genuinely honest at unlimited scale. We'd rather be genuinely useful to fewer people than nominally useful to everyone.
These principles are a commitment, not a description. They are what we hold ourselves to when a decision is difficult, when a feature is tempting, when growth would be the easier path.
If we ever move away from any of them, we'll say so and explain why. That's what transparency actually looks like: not a policy page, but a willingness to be held to what you said.
The full version of this lives at continio.app/beliefs.