Sciene - Lucas Lima

Prev. Case

Next Case

Sciene

Enterprise-grade AI solutions

My Role

Discovery, competitor analysis, low & high fidelity prototyping, usability testing & validation.

Year

2025

About the client

Sciene is an AI-first technology company that provides secure, customizable, enterprise-grade AI solutions.

Next project

Taking ownership of a complex platform, modernizing it, systematizing it, and scaling design operations inside a fast-paced AI company.

How it unfolded

Defining project scope

Defining project scope

Understanding users & workflows

Usability audit & discovery

Usability audit & discovery

Information Architecture redesign

Information Architecture redesign

Design System creation

High-fidelity prototyping

Launch & continuous improvement

The challenge

Revamping a complex AI platform required a continuous, multi-layered approach rather than a traditional linear design cycle. As the sole Product Designer, I partnered closely with AI Engineers, Developers, QA, Data Analytics, Sales, and CSM teams to modernize the entire experience, introduce scalable design foundations, and accelerate product delivery.

Below is how the process unfolded.

Platform Immersion & Usability Audit

I began by conducting a deep dive into the legacy platform to understand its structure, pain points, and technical constraints.

The audit uncovered critical issues:

Highly inconsistent components and outdated visuals
Fragmented interaction patterns
Overloaded screens with unclear hierarchy
Navigation friction and lack of discoverability
No design system or documentation
Accessibility gaps and alignment issues
Redundant flows created over time by different teams

This formed the baseline for a full-platform modernization roadmap.

I began by conducting a deep dive into the legacy platform to understand its structure, pain points, and technical constraints.

The audit uncovered critical issues:

Highly inconsistent components and outdated visuals
Fragmented interaction patterns
Overloaded screens with unclear hierarchy
Navigation friction and lack of discoverability
No design system or documentation
Accessibility gaps and alignment issues
Redundant flows created over time by different teams

This formed the baseline for a full-platform modernization roadmap.

Stakeholder Alignment & Strategy Mapping

I collaborated with PMs, AI Engineering, Sales, and CSM leadership to translate business goals into clear design objectives.

Together we defined:

The strategic role of each AI product (AIChat, Virtual Assistants, Insights Generator, Sales Performance Tool)

What “future-ready usability” means for an AI-first platform

The scope and sequencing of redesign efforts

Foundational design principles: clarity, scalability, performance, and security

Technical considerations like model response times, API output formatting, and knowledge ingestion workflows

This phase ensured that designer, engineers, and business teams were working toward the same North Star.

I collaborated with PMs, AI Engineering, Sales, and CSM leadership to translate business goals into clear design objectives.

Together we defined:

The strategic role of each AI product (AIChat, Virtual Assistants, Insights Generator, Sales Performance Tool)

What “future-ready usability” means for an AI-first platform

The scope and sequencing of redesign efforts

Foundational design principles: clarity, scalability, performance, and security

Technical considerations like model response times, API output formatting, and knowledge ingestion workflows

This phase ensured that designer, engineers, and business teams were working toward the same North Star.

User Research & Workflow Understanding

The objective

I conducted contextual interviews with customers, Sales reps, and CSMs to understand how different teams used the platform in real environments.

This research directly informed how flows and hierarchy should be restructured.

The key findings

Users relied heavily on AI tools but were slowed down by unclear inputs and non-standard UI patterns

Navigation contributed to repeated errors and support tickets

Users improvised external tools (Notion, Excel, manual documents) to fill product gaps

CSMs struggled to onboard clients due to the platform’s complexity

IA Redesign & System Architecture Foundations

Before touching high-fidelity UI, I redefined the information architecture to support a scalable future:

Clarified navigation groups and product boundaries
Introduced consistent patterns for multi-step workflows
Defined global rules for data hierarchy, table structures, and actions
Established clear entry/exit points for AI features
Standardized how AI results, errors, and loading states appear across products

This enabled a unified, predictable experience across all Sciene products.

Before touching high-fidelity UI, I redefined the information architecture to support a scalable future:

Clarified navigation groups and product boundaries
Introduced consistent patterns for multi-step workflows
Defined global rules for data hierarchy, table structures, and actions
Established clear entry/exit points for AI features
Standardized how AI results, errors, and loading states appear across products

This enabled a unified, predictable experience across all Sciene products.

Low-Fidelity Exploration & Flow Redesign

I redesigned every major flow in the platform, including:

AI onboarding
Knowledge ingestion & document upload pipeline
AIChat interactions & message handling
Real-time insights generation
Sales performance analytics
Role & user management
Error and edge-case handling
Settings, navigation, dashboards, and administrative tools

Multiple variations were tested internally to validate feasibility and clarity before moving into UI.

I redesigned every major flow in the platform, including:

AI onboarding
Knowledge ingestion & document upload pipeline
AIChat interactions & message handling
Real-time insights generation
Sales performance analytics
Role & user management
Error and edge-case handling
Settings, navigation, dashboards, and administrative tools

Multiple variations were tested internally to validate feasibility and clarity before moving into UI.

Building the New Design System (from scratch)

Since no system existed, I created a full-scale design system covering everything:

Color system, semantic tokens, elevation
Typography scale and rhythm
Grid, spacing, and layout foundations
Full component library (50+ components)
Interaction patterns
Data visualization guidelines
AI interaction rules (inputs, results, errors, tooltips, loaders, model states)
Accessibility guidelines and contrast compliance
Documentation for engineering handoff

This unified the visual language and drastically improved engineering speed.

Since no system existed, I created a full-scale design system covering everything:

Color system, semantic tokens, elevation
Typography scale and rhythm
Grid, spacing, and layout foundations
Full component library (50+ components)
Interaction patterns
Data visualization guidelines
AI interaction rules (inputs, results, errors, tooltips, loaders, model states)
Accessibility guidelines and contrast compliance
Documentation for engineering handoff

This unified the visual language and drastically improved engineering speed.

Validation & Iteration

Testing & Cross-Team Validation

Interactive prototypes were tested with:

Customers using real workflows
CSMs testing onboarding scripts
Sales validating demo scenarios
Engineers reviewing feasibility

Feedback loops were fast and iterative, ensuring designs solved real problems and aligned with technical reality.

Development Support & Handoff

To accelerate implementation, I delivered:

Detailed Figma specs
UX guidelines and component documentation
Accessibility and behavior notes
Flows annotated for engineering
Loom walkthroughs explaining rationale
Quick QA support during development

This significantly reduced back-and-forth and improved release speed.

Impact

Let’s work together.

Get in touch.

2024 Lucas Lima

Currently based in Curitiba - Brazil

Networking

Read.cv

More projects

Behance

Networking

Work

Samples

About

Contact