Last updated 5 August 2026
Virlia analyses public creator videos and returns a ranked shortlist with brand fit and brand safety scores, the evidence behind them, and draft outreach messages. It is a research and drafting tool. Every judgement it produces is a recommendation for a human to accept or reject.
Virlia is built on artificial intelligence, including third party large language models and vision models. Frame analysis, brand fit and brand safety scores, the written evidence attached to them, transcripts, summaries and every outreach draft are generated by AI. Your brief, your brand context and the material we analyse are transmitted to and processed by third party AI providers so that this can happen.
AI output is a probabilistic estimate, not a measurement and not a statement of verified fact. It can be wrong, incomplete, out of date, internally inconsistent, or different on a second run of the same input. It can also state something confidently that is simply not true, which the industry calls a hallucination.
In this product that can mean, without limitation: misreading what is in a video, misattributing a product, a claim or a sponsorship to a creator, transcribing speech incorrectly, misjudging tone or context including sarcasm and satire, missing a brand safety problem that is really there, flagging one that is not, scoring a creator who is a poor fit highly or a good fit poorly, misstating an audience, a follower count, a jurisdiction or a rate, and drafting an outreach message containing a factual error about the recipient.
Scores are opinions produced by software. They are not statements of fact about any creator, and they are not a background check, a legal opinion, a compliance certification, a credit or suitability assessment, or advice of any kind. Nothing Virlia outputs about a person should be treated as a verified finding about that person.
You must check the evidence before you act. Every score links to the video it came from precisely so that you can watch it yourself. You agree to verify AI output independently before contacting a creator, signing a contract, paying a fee, publishing anything, or making any other commercial, legal or reputational decision, and not to rely on Virlia as the sole basis for such a decision.
A human is always in the loop. Virlia recommends and drafts; you decide and you send. The product has no send capability by design. Virlia does not make automated decisions that produce legal or similarly significant effects on any person.
You are the sender and publisher of anything you send or publish from a Virlia draft, and you are responsible for its accuracy and for compliance with the law that applies to you, including advertising, endorsement disclosure, unfair commercial practice, anti spam and data protection rules.
All AI output is provided as is and as available, with no warranty of accuracy, completeness, currency or fitness for any purpose. If you find an error, tell us and we will look at it, but we do not promise any accuracy rate.
Your first 3 reports are free and we do not ask for a card to run them. After that you choose a plan.
Plans are billed monthly or annually in advance. Each plan includes a stated number of runs per month, which resets each billing period and does not roll over.
You can cancel at any time from the billing portal. Cancelling stops the next charge; it does not refund the period you are in. If we fail badly enough that you want your money back, ask and we will sort it out.
Prices can change. If they do, we will tell you before it affects a renewal.
We do not promise a campaign will perform, that a creator will reply, or that a shortlist is exhaustive. The product tells you what it saw in the videos it watched, and it says so when frame coverage is thin.
Model output can be wrong. Scores are judgements, not facts. Every score links to the video it came from precisely so you can check it before spending money on a partnership.
We depend on public platforms for video data. When a platform changes, coverage can change with it.
Do not use Virlia to harass creators, to send messages at a volume you would not defend, or to misrepresent who you are. The product deliberately has no send capability: it drafts, you send, and what you send is on you.
Do not resell the output as your own research product without telling your client that a tool produced it.
Do not attempt to extract our data beyond your own runs, or to reverse the analysis for a competing product.
Your brand context, briefs and runs belong to you. We use them to operate the product for you and for nothing else. We do not train models on them and we do not share them with other customers.
You are responsible for having the right to the brand information you enter.
We can suspend an account that breaks these terms, that abuses the free allowance across many accounts, or that puts the service at risk. Where it is reasonable to warn first, we will.
To the extent the law allows, our total liability is limited to what you paid us in the twelve months before the claim. We are not liable for lost profits or for the outcome of a campaign.
To the extent the law allows, we are also not liable for any error, omission, inaccuracy or hallucination in AI output, or for any loss, cost, wasted spend, contractual liability, regulatory penalty or reputational harm arising from your reliance on it, given the verification duty set out above. Nothing here limits liability that cannot be limited by law.
orhan@virlia.com.
Operating entity and governing law: to be confirmed before live payments are accepted.