Weaver-AI
Weaver-AI for schools, universities and research teams
A study protocol, a student record, a committee that has to be quorate. Each one has a rule attached, and the review arrives long after the meeting. Weaver-AI keeps the rules in view while the decision is being made.
Last updated 3 September 2026
Why this work needs care
A research team designs a study, collects data and publishes. Somewhere between those, somebody was supposed to check that the approved protocol still covers what is actually being done. The person best placed to notice is the one running the experiment.
Student records sit under rules most staff have read once. The question is never whether the rule exists. It is whether anyone remembered it during the twenty minutes the decision took.
A committee reaches a decision and the minutes record the outcome. What is missing a year later is who was present, what they were working to, and on what basis.
How Weaver-AI helps
Decide in a room that knows the rules
Open a decision room for the question. An ethics submission, a data-sharing agreement, a student disciplinary matter. Say which standards apply: the Common Rule for human-subjects research, FERPA for education records, GDPR where participants are in the EU. Weaver-AI names them for every member and says which one bears on what is being discussed.
Answer from your approved protocol, not the literature
Add the approved protocol, the consent forms and your institution's own policies to the workspace, and ask against those. Weaver-AI says what it could not find rather than answering from research practice in general. Research practice in general is not the protocol your ethics board approved.
Write up the committee that decided it
An ethics panel, an admissions meeting or a thesis defence gets notes, the decision, and who owns what next. It also gets an attendance record of who was actually present. Where the composition of the committee is itself a requirement, that record is the evidence it was met.
Keep it as long as the funder requires, and no longer
Research records outlive the project and student records outlive the student. The periods differ by funder, by jurisdiction and by subject. Retention is set per account in whole years, with a legal hold for the obligations that have no end date at all.
What makes this different from a general chatbot
Weaver-AI is not your ethics board and does not grant approval. It identifies rules that may apply and says to check them, while the study design can still change.
A room says what it is working to. 45 CFR 46, FERPA, your own research code. Every member can see that list, so nobody has to trust that somebody checked.
Education records and research data are among the most restricted material an institution holds. Access is scoped to the workspace, and Weaver-AI answers from what that workspace was given rather than from the open web.
Meeting attendance is disclosed and recorded. A committee that has to be constituted a particular way leaves evidence that it was, rather than an assurance that it was.
It can use approved documents, meetings you invited it to, and applications you connected. It also says what it could not verify.
What Weaver-AI will not do
It joins no meeting you have not turned on, reads no calendar you have not connected, and is announced in every meeting it attends. Meeting audio is never retained. Transcripts are held for thirty days and then deleted. The notes written from a meeting are kept for as long as the account owner sets, and the default is ninety days. Credentials spoken aloud are removed before anything is stored.
For the full account, see the Privacy Policy. To get started, open app.weaversec.com, or read the documentation first.