Better context.
Human judgment.
A memory for later.
Libelle helps a person think. It doesn't make the decision for them.
Someone tells us about themselves. Libelle helps bring that evidence into the context of what TCUS is doing. A human considers what might be worth exploring — and keeps track of what happens next.
From introduction to a thoughtful next step.
The system is designed around a simple sequence. Each part has a different job.
You share.
Background, experience, interests, curiosity, availability, résumé, and other useful evidence.
Libelle helps organize.
It brings together what the person actually shared with the context of what TCUS is doing now.
A human considers.
Someone at TCUS decides what seems worth discussing, what remains unclear, and what should happen next.
Libelle remembers.
The takeaway, decision, next action, and a reason to return later don't have to disappear.
What you said should stay different from what the system thinks.
Libelle deliberately keeps source evidence, generated interpretation, and human judgment from quietly becoming the same thing.
What the volunteer actually shared.
Their own answers, résumé or CV, submitted links, availability, interests, and other information they intentionally provided.
Libelle makes the evidence easier to work with.
Parsed résumé information and structured views can help a reviewer see the relevant evidence without repeatedly reconstructing it from scratch.
AI can help the reviewer interpret what they see.
It can summarize what appears relevant, point out things worth noticing, suggest possibilities to explore, and identify important unknowns.
The reviewer decides what actually matters.
A person at TCUS can record their own takeaway after considering the evidence and any assisted interpretation.
What are we actually doing?
Reach out. Wait. Revisit later. Close the loop. Start a conversation. The operational state remains an explicit human action.
Helpful is different from authoritative.
Libelle can use AI to assist understanding. That does not give the model authority over the person, the decision, or the relationship.
AI can help with
Summarizing relevant evidence.
Helping a reviewer understand unfamiliar experience.
Surfacing things that may be worth noticing.
Suggesting possible things to discuss or explore.
Making important unknowns visible.
AI does not
Rank volunteers.
Decide whether someone is accepted or rejected.
Assign someone to a project.
Turn a possibility into a commitment.
Replace the conversation between people.
More information isn't automatically better.
The current design is intentionally bounded. Libelle should use information that is relevant to understanding a possible contribution — not quietly gather everything it can find about a person.
Useful context
Not automatically part of the picture
The next step is a conversation, not an assignment.
A possibility is simply something that may be worth exploring. People still have to decide together whether it makes sense.
Something starts.
The volunteer and TCUS find a contribution that both sides actually want to pursue.
Not yet.
There may be no useful or responsible next step right now. Libelle can preserve why — and when it might be worth looking again.
Enough to pick up where you left off.
Libelle doesn't need to remember every interaction. It needs to preserve the consequential parts so another reviewer — or the same reviewer months later — can understand what happened without reconstructing the entire history.
Start with what you bring.
You don't need to know what role you fit. Tell us about yourself and give us enough context to start thinking with you.