Trust
Built so you can be honest about your team
A useful conversation about a team involves saying things you would never put in an email — that someone is stretched, that a colleague is in the way, that you're not sure a strong performer is actually the problem. That's the risk this page is about, and it's the one we designed for first.
What we need to know about your company
Not your company name. It's optional, nothing asks twice, and Ygora works the same either way. What shapes the suggestions is the context: what you lead, roughly how big the company is, who owns it, and what you have been asked to deliver. All of that is optional too. A team of ten needs something quite different from a team of five thousand, which is why we ask about the structure rather than the identity.
If you do give the name, we use public sources — the Norwegian company register and similar — to sharpen the guidance. That's your choice, not a requirement.
We don't connect to your systems. No HR system integration, no calendar access, no mailbox access, no document upload. Ygora reasons over what you choose to tell it, which means the amount of company data at risk is bounded by what you typed.
What happens to what you say about a named person
This is the question that matters, so here are the mechanisms rather than reassurances.
- Ygora's own emails never carry what you said. They are a doorway, not a briefing: they say something is waiting and link to it. No plan text, no names of people, no figures from your plan, and nothing in the subject line, because a subject line shows on a lock screen. This is enforced in the code, not by policy: the functions that send those emails accept nothing from your plan or your team. The most one of them takes is your own first name, and one takes a count: the one email you can ask for when three people have answered a team link says how many have answered, never what they said.
Two things you do yourself are different, and you see both before they go. An invite is written in your own email program, so we never send it. A note to someone in the pool is sent by Ygora on your behalf, and that person stays anonymous to you until they answer. You write it, you can change every word, and it goes only when you press send. The note we draft does not name your company, and that is yours to change too. - It is never reused for another leader. We do not cache or share answers between customers. A reply is specific to one team; serving it to someone else would be both useless and a leak.
- It is never used to train a model. Not ours, and not our providers'.
- A plan holds no names. The work, the roles and the decisions are about roles and teams, and Ygora never asks you to name someone on your team. The bench is the one place a name lives: people you invite or add yourself, and people who answer an invite. To learn which kinds of suggestion actually help, we keep a short record each time you confirm, change or dismiss one: the kind of suggestion, what you did, and a little fixed context (what you lead, rough size bands, timeframe). It has no wording in it, and never your company name, your goal or your notes. If you choose to share the wording of the suggestions as well, in Settings, we keep that too, with emails, links and phone numbers taken out. That record is not your plan, it is never used to contact you or to rank anyone, it is deleted with your account, and you can switch either part off in Settings.
- The specifics stay behind sign-in. Your plans, what is written from them, your conversation with Ygora and your bench are visible only to you when you're logged in. A colleague at the same company does not see your workspace. The conversation is saved with the plan, and you can delete it.
- People in the pool chose to be there. They set their profile to visible to companies and say what they want to do next. They stay anonymous to you until they answer a note.
One thing we ask of you in return: don't enter health information, union membership, or anything else in the special-category bracket about a team member. Ygora doesn't need it to be useful, and we'd rather it never arrived.
When you run a team check-in
You can send your team a link and ask for their feedback on what steals their time. This is exactly what that does.
- What your team sees. The names of the pieces of work on your list and the titles of the roles on your team. Nothing else: not your goal, your plan, your notes, your charter, or what anyone else has said. They need no account and do not sign in.
- What your team can say. For each piece of work: that it steals their time, that AI could support them in it, or that it should stay with people, with a few words if they like. They can pick their role from your list or leave it out. The form asks them to write about the work, not about people. To get them started it suggests a few ideas for their role, written by Ygora's AI from the role title and the names of the pieces of work. They can also just tell Ygora in their own words, in a short chat that helps them make an idea specific, say whether it can start within the team or needs another team, and saves it as a note.
- What we store. The note, the piece of work, which of the three it was, the role if they gave one, and when. For an idea from the chat, also whether it can start within the team or needs another team, and which one (never a person). We do not store a name, an email address, an account, an IP address or device details with a note. Our host keeps ordinary server logs, as any website does; they are not part of what you see and are not tied to a note.
- Who sees it. Only you, grouped by role, on a board in your plan. You can remove any note and turn the link off at any time. The AI and the chat: what your team taps or types in the form is not read by an AI model. If someone uses the chat, what they type there is sent to Ygora's AI provider (Anthropic, United States, not used for training) to reply and to pick out the ideas it saves. The conversation itself is not kept, only the notes it saves, and the page tells them so before they type.
- An honest limit. There is no minimum number of answers, so in a small team a role can point to one person. The form tells people to leave the role out if they would rather not be recognized. Anyone with the link can add notes, so share it only with your team and turn it off when you are done.
- One email, if you ask for it. When you make the link you can tick a box to be emailed once when three people have answered. It goes only to you, says how many have answered and links to your Home page. It never says what anyone wrote, and the people who answer are never emailed.
- What we measure. Only counts: how many links are made, how many notes come in, whether through the chat or a tap, and how many ideas are used in a plan. Never what anyone wrote, a role name, or anything that identifies a person.
- It goes when the plan goes. The link and the notes are deleted with the plan, and with your account.
Tell your team why you are asking before you send the link. Where you have employee representatives, bring them in first. Ygora is not a way to judge anyone: it asks about work, never about people.
Who processes it, and where
Encrypted in transit and at rest. A short, named list of providers, and no others:
- Supabase — database and sign-in, hosted in the EU.
- Anthropic — the model behind the suggestions and the chat. It proposes the work, who does it, who decides, the lines of your charter and work descriptions and the notes on your bench, and it writes down what you tell it in the chat bar. You confirm or change every one. It also powers the optional chat on the team link. It never produces the findings or the list of what to do: those come from fixed rules over what you have confirmed. Processed in the United States. Not used for training.
- Vercel — application hosting. United States.
- Resend — delivery of sign-in and service emails, which carry none of what you or your team said.
Where data goes outside the EU/EEA we rely on the EU Standard Contractual Clauses. We do not sell personal data, we run no advertising or third-party tracking cookies, and we share nothing beyond the providers above except where the law requires it. Ygora AS is the data controller throughout; the full legal detail, including your GDPR rights and how to exercise them, is in the privacy policy.
Getting your data out. Download everything, or delete the account outright, from Settings — you do it yourself, in the product, and neither goes through us. No support queue, no retention offer. If you would rather ask a person, hello@ygora.ai reaches a founder.
What we don't have yet
We're a small Norwegian company and we'd rather tell you this than have you find out in a procurement form. Ygora does not currently hold a SOC 2 report or ISO/IEC 27001 certification. We haven't started an audit, and we won't claim a badge we don't have.
What we do run is on this page: data minimisation by design, a short named subprocessor list, encryption in transit and at rest, multi-factor access to production, per-leader isolation, and the structural limits on email and reuse described above. If your organisation needs a formal audit before you can buy, tell us — it moves up the list when a real customer needs it, and we'll be straight with you about the timeline rather than promising a quarter.
We'll sign a data processing agreement, answer a security questionnaire, and talk to your IT or legal team directly. Email hello@ygora.ai and it reaches a founder, not a queue.
Questions we haven't answered here go straight to us.
Email hello@ygora.ai