All results
CarRentRightAI onboarding
Replaced a 30-field form with one conversation, without losing the facts a legal case needs
A chat-first intake that turns a user's story into a structured, auditable claim.

30 → 1
form fields replaced by a chat
4
storyline blocks guide the AI
1,000
cases/month target without manual review
logicflow · carrentright · the work
before
The challenge
- People fighting unfair charges have a story, not a dataset. Long forms put the load in the wrong place.
- Missing facts at intake become very expensive later, because the founder evaluates each claim by probability.
- The founder could not scale past about 1,000 cases a month while evaluating intake by hand.
what we built
The solution
- A chat-first 'Start Your Case' wizard: the user writes freely, AI turns it into a structured claim.
- The original story is stored word for word next to the AI summary, for an auditable record.
- A four-block storyline framework drives the AI's follow-up questions.
- Server-side AI calls validated against the data model.
after
The results
- Achieved conversational intake across all V1 user journeys, without a 30-field form.
- Achieved structured claims ready for evaluation, without the founder doing manual intake.
- Achieved a legally auditable record (story, AI summary, structured claim), without losing the user's own words.
“Super simple, super clear. No fluff.”
built with
- Next.js
- NestJS
- Postgres
- LLM
- AWS
logicflow · your build

Thinking about your own build?
Get a clickable prototype in 48hFree. 2 prototypes a month. Yours to keep.

