All work
Legal-tech2025

SciJustice. Bringing safe, structured expert consultation into judicial workflows.

Many legal decisions need domain experts. Courts often need specialists to support complex cases. SciJustice modernises this through a secure digital platform built on four design goals: faster expert matching, compliance at every step, secure communication, and efficient management across multiple active cases.

Role
UX/UI Designer
Duration
8 weeks
Platform
Web platform
scijustice · screens
01Judge dashboard

At-a-glance overview + clear next best action

02Inquiry details

Natural-language intake the AI can later structure

03Expertise selection

Field, subfield and time-commitment filters

04Compliance checkpoint

Boundaries reviewed before any request goes out

05Generated request review

AI summarises the inputs into a structured request profile

06Expert match results

Match scores explain why each expert was recommended

07Request sent · locked session

Affidavit-gated session before any chat begins

08Expert dashboard

Pending requests, ongoing chats and summary tasks in one view

09Consultation request review

Conflict-of-interest reminder, request details, accept / reject / hold

Project Overview

The context.

  • 01

    Many legal decisions require domain experts.

  • 02

    Courts often need specialists such as doctors, engineers, financial analysts or forensic professionals to support complex cases.

  • 03

    SciJustice was designed to modernize this process through a secure, transparent digital platform.

Design goals

What we set out to achieve.

01

Faster Expert Matching

Help judges find the right expert quickly — match score, credentials and availability surfaced together.

02

Compliance

Ensure compliance at every step with built-in safeguards and explicit checkpoints before sensitive actions.

03

Secure Communication

Enable trusted conversations and document exchange inside affidavit-gated sessions.

04

Efficient Management

Simplify coordination across multiple active cases with clear status and a single dashboard.

Process

How I worked through it.

01

Dashboard for the judge

An at-a-glance overview of ongoing requests, completed consultations and expert activity. Designed to reduce cognitive load, prioritising the next best action — start a new expert request — over information density.

02

Guided request creation

A 4-step wizard: inquiry details (natural-language intake the AI later structures), expertise selection (field, subfield, time commitment), compliance checkpoint, and a review screen with an AI summary the judge can edit before submitting.

03

Expert discovery & selection

Match results explain why each expert was recommended — match score, credentials, availability — instead of dumping random profiles. CV details open in a focused modal so judges can validate experience without losing context.

04

Secure communication & finalisation

Locked sessions with required affidavits before chat begins, a structured consultation chat that reinforces boundaries around case-specific information, and a session summary with downloadable records to create a reliable audit trail.

Outcome

The outcome.

01
Step-based wizard approach
02
Explicit compliance checkpoints
03
Clear status indicators
04
Neutral, minimalist design
Learnings

What I’d carry into the next one.

  • In high-stakes systems, clarity isn’t a convenience — it’s essential.

  • Trust signals (status, compliance, attribution) are first-class UI, not microcopy.

  • Step-based wizards beat single long forms when the cost of a wrong submission is high.