JuriCloud
JuriCloud
Designing a conversational legal AI workflow around trust, source grounding, and reusable systems
Designing a conversational legal AI workflow around trust, source grounding, and reusable systems
Summary
Summary
I designed a legal AI research workspace for JuriCloud that helped users move from open-ended legal questions to source-aware, verifiable answers. The product supported prompt-based research, file and case attachments, citation-backed responses, and reusable design system patterns for complex conversational states.
I designed a legal AI research workspace for JuriCloud that helped users move from open-ended legal questions to source-aware, verifiable answers. The product supported prompt-based research, file and case attachments, citation-backed responses, and reusable design system patterns for complex conversational states.
Scope
Scope
Design System
Conversational AI Design
Design System
Conversational AI Design
Client
Client
JuriCloud
JuriCloud
Context
Context
The Problem
The Problem
Legal professionals do not trust AI just because it sounds correct. They need to verify where an answer came from, what source supports it, and whether the reasoning can hold up under review.
Legal professionals do not trust AI just because it sounds correct. They need to verify where an answer came from, what source supports it, and whether the reasoning can hold up under review.
Context before generation
Citations inside the response
Source verification without breaking flow
Clear separation between generated text and source material
Most chat-based AI interfaces optimize for fluency.
Legal research requires something else
What I built
The goal was not just to make the app work. It was to make the automation feel dependable enough that users could stop thinking about it.
I designed and built the product end to end, including:
the menu bar interaction model
setup and settings flows
folder monitoring behavior
local conversion pipeline
processed-file tracking
safe output handling for duplicate file names
What I designed
A reusable design system for JuriCloud’s legal AI workflows, covering conversational research, source inspection, matter context, prompt templates, attachments, loading states, citations, and errors.
We started with color.
The palette is warm, restrained, and evidence-forward. Neutrals carry most of the interface, gold is reserved for primary actions and active evidence states, and semantic colors are used only when they clarify risk, success, or status.
Typography for legal reading.
The type system is designed for long-form legal answers, compact controls, and dense source metadata. Headings guide scanning, body text supports sustained reading, and labels keep filters, citations, and status states legible at small sizes.
Components built around conversational workflows.
Components were not created as isolated UI parts. Each component supports a specific AI moment: asking, scoping, attaching, searching, verifying, citing, saving, and recovering from uncertainty.
I designed the workflow across four connected layers:
This made the product feel less like a generic chatbot and more like a legal research workspace.
Research entry
Empty and returning-user states for starting or resuming research
Prompt system
Support for short questions, long factual scenarios, templates, and refinement
Context building
File uploads and attached cases to ground the model before generation
Answer review
Readable responses with inline citations and verification patterns
The design system made the workflow scalable
The design system made the workflow scalable
Because the product spanned prompt entry, attachments, answers, and verification, I designed it as a reusable system rather than a sequence of screens.
Prompt composer variants
File and case chips
Template selectors
Response blocks
Citation badges
Empty, active, and loading states
The system include
This kept the workflow consistent as complexity increased and made the experience easier to scale.
This kept the workflow consistent as complexity increased and made the experience easier to scale.
Reflection
Reflection
The biggest lesson from this project was that trustworthy AI interfaces are not defined by polish alone. They are defined by how clearly they show context, evidence, and boundaries.
The biggest lesson from this project was that trustworthy AI interfaces are not defined by polish alone. They are defined by how clearly they show context, evidence, and boundaries.
