UXplorer 2026 | Agent Design Challenge

UXplorer’26

Reimagine

The worlds first design led agent design challenge

Agents are being deployed everywhere. On top of broken processes. Inside inefficient workflows. Automating the mess instead of fixing it. We've seen this before. We called it digital transformation. And most people got it wrong. The ones who get agentic AI right won't just build agents. They'll reimagine the systems those agents live in. From the ground up.

That's what UXplorer 2026 is looking for.

Submit Your Concept Note

The

Challenge

Most agents being built today are going to fail.Not technically. They'll run fine. Process requests. Generate outputs.But nobody will trust them. Nobody will use them. And in 12 months someone is going to ask why we spent all that money.Because deploying an agent on top of a broken process just gives you a faster broken process.

The question was never: can we build an agent that does this task.

The question is: should this task exist at all? What does the system look like if you reimagine it from the ground up not just deploy into it?

Almost nobody in the agentic space is asking that.

UXplorer 2026 is built for the ones who will.

Three things that

Make this different

Reimagine,
dont just automate

We're not looking for agents that replicate existing workflows. We're looking for entries that challenge why those workflows exist. The best concept notes will throw out the current process entirely and ask what should this system look like if we designed it from scratch?

Real domains,
real stakes

Pick from industries where AI disruption is highest finance, healthcare, legal, customer service, operations backed by Anthropic's labor market research. Identify problems with genuine market potential and measurable business ROI. Not hypothetical. Not academic. Real.

Concept to deployment,
with Mox-F

Shortlisted participants get trained on YOXA and our proprietary Multi Agent Orchestration Experience Framework MOX-F. The framework we built at Yuj while developing our own agentic product. The one that taught us the difference between an agent that runs and an agent that changes business outcomes.

How

It works

Concept note

Due 24th Jul

Pick your domain. Reimagine the system. Define the problem worth solving. Show the ROI. One page. Evaluated on problem clarity, system reimagination, market viability, and originality.

YOXA Training

6th-17th Aug

Top 20 shortlisted participants get trained on YOXA Yuj's proprietary multi-agent orchestration platform and on MOX-F, the framework built to reimagine agentic systems before a single agent is deployed. 10 days. Hands-on. Real tools.

Build & Display

20th-31st Aug

Take your reimagined system to a working prototype. Present to a jury of enterprise design leaders and agent architects. Top 3 win.

What

You win

The real prize isn't the cash. It's what you walk away knowing and who you walk away knowing it with.

Training
on MOX-F

Yuj’s proprietary framework built from real-world agentic product development. Exact system used in enterprise-grade agent deployments.

Hands-on access
to YOXA

Yuj's own platform. You don't just learn the framework you build on the platform built around it.

YOXA
Token Pool

Access to tokens to keep building after the challenge ends. Your learning doesn't stop when the challenge does.

Agent Designer
Mentorship

Hands-on, targeted mentorship for winners, with practical guidance tailored to the agentic platforms they’re building.

Co-creation with
Yuj's AI Agent Team

The top 3 winners earn the opportunity to collaborate and co-build real-world AI agentic platforms together.

Up to $2,000 USD
in Cash Prizes

Distributed across the top 3 winners because exceptional work deserves meaningful recognition and reward.

Who

Should Apply?

You think in systems.
You care about outcomes, not just outputs.
You’re curious about AI. not as a shortcut but as a design material that can fundamentally change how work happens.

You don't have to be a designer or an engineer. You have to be willing to throw out the current process and reimagine it entirely.

That combination of system thinking, design judgement, domain curioisity is what we call an Agent Designer.

Right now, the world doesn’t have enough of them.

Individual/ group entries allowed. Open globally.

What makes a

Winning entry

Most entries will automate an existing workflow. Winning entries will question whether that workflow should exist.

The best concept notes will:

Identify a high-impact domain where AI displacement creates a genuine reimagination opportunity.
Challenge the current process not just automate it.
Demonstrate system thinking: what the agent decides, acts on, and hands off to humans.

Define an agent system designed from the ground up not deployed on top of what's broken.
Show measurable business ROI for an enterprise or business buyer.
Prove it's buildable, sellable, and scalable.

Concept note

Guidelines

One page. One reimagined system. Evaluated on:

Domain
& Problem

Which AI-disrupted industry? What specific broken process are you reimagining not just automating?

Target
User

B2B, B2C, or enterprise? Who uses this, in what context, and what does their current broken reality look like?

Agent
Concept

What does it do autonomously? Where does it defer to humans? What in the current workflow does it eliminate entirely?

System
Thinking

How do components interact? What data flows through? What decisions does the agent make versus what it escalates?

Technical
Architecture

What tools, frameworks, or APIs power your agent? How does data flow through it? Show how it’s built and deployed.

Market
Viability

Who buys this? What's the ROI? What does the enterprise pay for and why? What's the market size?