Amara
Amara, 18. In extended foster care. First semester at Valencia College, Orlando.
This is an illustrative account. Every detail is invented.
Watch
Full case
Situation
Amara is 18, in extended foster care, and in her first semester at Valencia College in Orlando. She has a stable placement and a plan. What she doesn't have is any idea that the state has money set aside for someone exactly where she is.
Why this case matters
This is the doing-well case. It answers the objection that the youth who are fine don't need this. Amara is fine, and she may still be leaving thousands of dollars a year on the table, because eligibility is scattered across six programs, four agencies and a caseworker who has forty other kids. The product's job here isn't rescue. It's turning eligibility into money in hand.
The feature she used
Benefits. From one assessment and her profile, the product found eight programs that may fit her today, and it says so at the top of the screen. Three are the headline: PESS, a monthly stipend while she's in school. ETV, up to $5,000 a year. And the tuition exemption, which waives tuition and fees at Florida public colleges. Under each card is a sentence that starts "You're seeing this because", so she can read why.
What changed
The documents PESS asks for were already in her Vault, with a plain-language note on which programs each one unlocks. Her Journey Record shows what she did and when: she gathered her FAFSA confirmation and her placement letter, added both documents, and submitted her PESS application. Each entry is dated, and none of it can be edited after the fact.
Said out loud
Amara didn't need rescuing. She needed someone to show her the eight programs that may be hers, in one place, with the reason written under each one. The product did that on her first day, from an assessment and a profile.
The federal line
Every dollar on Amara's screen is already appropriated. PESS and ETV are funded through the Chafee program, federal money Florida already receives and already reports on. The state's problem isn't finding funds for her. It's that the funds don't reach her, because eligibility lives in six programs and she doesn't know they exist. ACF measures exactly this. NYTD surveys youth in care at 17 and follows up at 19 and 21, and two of its outcome elements are whether she's enrolled in school and whether she received educational aid. Amara's screen lines up with both.
Honest note
These programs match today because Amara's profile carries her care history. A youth who only completes the assessment, without a profile, will not see all of them yet. That capture gap is the next build.
How her information is protected
- She owns it. U Navigate is not the state's system of record. Her account is hers. Nothing about her is shared with a caseworker or anyone else unless she chooses to share it.
- The record can't be rewritten. Her Journey Record is append-only. Entries can be added. They cannot be edited or deleted after the fact, by her or by us.
- Every row is locked to her. Database access is enforced row by row. A request for her data returns her data only, even if our own code has a bug.
- What the AI is given, exactly. Lexi knows her first name and her state, so she doesn't have to repeat herself. Her last name, email, phone number, address, date of birth and account identifier are never sent to any AI provider from her account, and an automated test fails our build if that changes. A document she chooses to upload is different, and the next line says how.
- When she uploads a document, the AI reads it. That's how Lexi summarizes a lease or finds a deadline on a form. The document is hers, the summary is hers, and neither is used to train anything.
- No ads. No data sales. Ever.
- What isn't done yet. We don't yet have a Business Associate Agreement or a zero-data-retention commitment from our AI provider. Until we do, we tell agencies that plainly and don't accept HIPAA-covered data.
Engineered by co-founder Shane Cupid.