Velocity AI is an AI-powered forecasting platform built for fashion supply chains. It helps planners upload data, interact with an AI system that understands their business context, run forecasts, identify risks, ask why outcomes are changing, and take decisions back into their workflows.

When I joined BlueKaktus in February 2025, Velocity AI was not yet a product. It existed as an idea, a few early discussions, and a strong belief that forecasting in fashion could be fundamentally better if AI was designed around real planning workflows.

This document captures how Velocity AI moved from that idea to a working MVP and a clear future direction. It is written as a hybrid of a UX case study and product documentation, because that reflects how the work actually happened.

<aside> 💡

TL;DR

• Product: Velocity AI, an AI-native forecasting platform for fashion supply chains
• Problem: Forecasting is fragmented, slow, and hard to trust in fast-moving fashion environments
• My role: Founding designer shaping UX, product structure, and interaction models from idea to MVP
• What makes this hard: Multi-agent AI systems, long-running tasks, explainability, and enterprise trust
• Outcome: A working MVP with a clear path toward control towers and decision-centric planning workflows </aside>

Velocity AI MVP Flow

https://vimeo.com/1145836975?share=copy&fl=sv&fe=ci

Design Brief & Future Vision Flow

https://vimeo.com/1145836863?share=copy&fl=sv&fe=ci

Inventory PO Release Flow

https://vimeo.com/1145836927?share=copy&fl=sv&fe=ci