Hullo
-->
Back to Portfolio
Milvian Group

MilvianAI

An enterprise-grade Gen-AI collaboration platform that enables teams to build, deploy, and manage AI-powered workflows—from data ingestion to intelligent automation.

Enterprise AI Gen-AI Platform Collaboration UX Strategy

ROLE/TEAM

Lead Product Designer
UX Designer
UI Designer
UX Writer
IA

TOOLS

Figma/Figjam
Nova
Adobe CC
Jira
SharePoint
Lovable
Pendo

DATE/DURATION

9 months
Jan 2025 — Sept 2025

Situation

Milvian Group pinpointed Generative AI as the next frontier — and the cornerstone of its next-generation platform. MilvianAI isn't just another tool; it's the intelligent engine designed to unify and amplify the entire Milvian ecosystem.

Think of it as the central nervous system for AI-powered enterprises — seamlessly connecting Aqueduct (smart, data-driven facility management) and Tiber (legacy-to-digital transformation for industrial & manufacturing data) into one cohesive, future-ready suite.

"The Operating System for AI-Driven Teams"

As a concept, MilvianAI was designed to be an enterprise-grade AI collaboration platform where teams don't just use AI — they build with it, deploy at scale, and manage intelligent workflows end-to-end, from raw data ingestion to autonomous decision-making. It fuses conversational AI, no-code automation, real-time collaboration, and enterprise-grade security — all within a unified, intuitive workspace.

Offsite Design Sprint

As a first step, the team convened in Houston for a week-long in-person offsite that involved all constituent groups (Executive Leadership, Product, Marketing, Engineering, Sales, etc).

Over the week-long session, we engaged in discovery and design exercises to help define the requirements for the platform and validate ideas and concepts that we saw as key features and capabilities for the app.

design sprint workshop team collaboration whiteboard session sticky notes enterprise AI platform planning

Sprint Outcomes

  • Platform requirements definition
  • Key feature validation
  • User journey mapping
  • Stakeholder alignment
  • Technical feasibility assessment
  • Go-to-market strategy

Competitive Research

Key Findings

As part of some of the exercises, we performed an analysis of leading AI platforms, as well as newer ones that offered unique capabilities and features. From that exploration of competitive platforms, we identified some areas to address as we began to create MilvianAI.

Role-Based Access Issues

Confusion around role-based access and team structures

UI Complexity

Overwhelming UI with too many entry points

Feedback Gap

Lack of feedback after actions (e.g., "Did my automation save?")

Authentication Friction

Friction in authentication — especially for enterprise SSO users

Onboarding Issues

No clear onboarding path for non-technical users

Poor Discoverability

Poor discoverability of integration functionality and workflow automation tools

Task

As Lead UX, UI Designer & Research Lead, as we left the offsite, I was tasked with delivering a unified, human-centered UX strategy across the entire MilvianAI product lifecycle, resulting in the following deliverables:

User Personas

Representing diverse roles, industries, and tech proficiencies, refined from discovery exercises and market research

Information Architecture

Scalable, role-aware navigation schema for the app and its features

Chat UI

Conversational interface for prompting, debugging, and iterating AI models

Feedback Mechanisms

Micro-interactions, success/error states, undo patterns

Login & Signup Flows

Frictionless SSO + social authentication with progressive profiling

Droplet Studio

Simplified + Advanced views for workflow building with live previews

Teams & Groups UX

Intuitive creation, switching, and permissions management

Data & App Integration

OAuth flows, universal MCP connectors for broad tool accessibility

Workflow Automation

Drag-and-drop builder with AI-assisted suggestions

Marketing Strategy

Support for blogs, sales outreach, and digital marketing vehicles

All deliverables were validated through internal testing, market research, analytics, and stakeholder alignment.

Visual Design & Guide

Design Approach: Started with a defined array of fonts, colors, and styles from a new expanded brand guide I was also tasked with creating. These served as inspiration for all proposed solutions and designs.

As a shift, for the beta, the colour palette was minimized to use mostly grays with an eye towards refinement for the final v.1 release.

Style Guide

minimalist design system style guide grayscale color palette typography components UI elements modern clean interface

Action

We began by hearing users—discovering frustrations and core needs. From existing customers, friends, family members our foundation: people-first decisions, simplicity breakthrough, confidence where once stood confusion. We brought it all to life.

Research & Analysis

  • User interviews across roles and industries to target and prioritize higher value features
  • Analyzed behavioral data from Hotjar, Pendo and anecdotal data
  • Created detailed user personas with goals and pain points
  • Mapped IA using card sorting and tree testing
  • Redesigned Gen Chat UI with split view and AI explanations
  • Implemented micro-feedback (toasts, undo, loading states)
  • Streamlined login flow with SSO auto-detection
  • Simplified team management with visual workspace switching

UI Screens

enterprise AI platform dashboard interface chat UI workflow automation screens grayscale design modern clean

Private Beta & Marketing Site

private beta testing interface marketing website landing page AI platform clean modern design responsive mobile ready

Deliverables

Key deliverables created for this effort include:

Offsite Exercises

Figjam boards and followup summary decks

Sitemap & Wireframes

High-fidelity wireframes and site architecture

Gen Chat UI Specs

UI files with mobile optimized view planned for post v.1

Login/Signup Flows

Complete authentication user experience

Onboarding & Help UX

To be implemented on a Pendo back-end

Orgs and Teams UX

Complete team management flows

Integration UX Pattern

Comprehensive integration pattern library

IA Diagrams

View mindmap

Visual Guidelines

Brand ethos translated into digital experience

Marketing Site

Mobile ready site

Result

By centering UX around real people, we turned a complex AI platform into an intuitive, empowering experience that drives adoption, trust, and results. We also were able to create novel solutions to some of the common pitfalls, like overly complex screens and confusing user flows for key features like workflow automations.

Platform Launch
Successfully launched enterprise AI platform
User Adoption
Intuitive design drove rapid user engagement
Innovation
Novel solutions to common AI UX challenges

Key Achievements

  • Simplified complex AI workflows into intuitive interfaces
  • Created scalable design system for enterprise AI
  • Established clear onboarding paths for all user types
  • Reduced authentication friction with smart SSO detection
  • Built comprehensive feedback systems
  • Unified Milvian ecosystem into cohesive platform