AI Trading. Human Control.

A design concept for AI-assisted trading. Figures in the interactive prototype are illustrative; it does not execute trades.

My role Principal Product Designer

  • Agentic AI
  • FinTech UX
Explore the product Open full prototype in a new tab
NeuroTrade portfolio dashboard and mobile trading experience displayed on a tablet and phone
Principal Product DesignerDesign concept · interactive prototypeFigma · FigJam · Framer

MY CONTRIBUTION

Making proposed automation inspectable.

I led product and UX direction, mapped account access and recovery, and designed the portfolio, trading, bot-setup, and monitoring screens. I built the interactive prototype and supporting FigJam flows and data model. Live exchange connections, model inference, and trade execution are not established by these artifacts.

PRODUCT EXPERIENCE

From insight to action.

Explore the product screens. Select a view to inspect the original design at a useful size.

Choose a screenSelect to update the preview.
SCREEN 01 / 08

Portfolio dashboard

The landing dashboard combines balances, asset holdings, market context, and agent activity so people can orient before choosing a next action.

Figures shown are sample interface data, not verified investment results.

AI WORKFLOW

SECONDARY INTERACTION / AI & HUMAN OVERSIGHT

Automation should be understandable at every step.

I designed the bot flow as a four-step sequence: connect, configure, activate, and monitor. The screens represent proposed automation behavior in a prototype, not a live trading system.

STEP 01

Connect

Select an exchange and establish the account connection before bot setup. The flow makes account access a distinct prerequisite.

Open the interactive prototype in a new tab

THE CHALLENGE

Complex trading. Clearer decisions.

Trading products ask people to interpret dense market information, understand automated strategies, and stay in control while making consequential choices.

  • Make portfolio context easier to read.
  • Make automated strategy choices visible.
  • Keep control clear before and after setup.
The landing dashboard combines balances, asset holdings, market context, and agent activity so people can orient before choosing a next action. Figures shown are sample interface data.

KEY DESIGN DECISIONS

Making complex automation understandable.

Three design choices connect the challenge to the interface: orient first, expose setup decisions, and keep configured automation inspectable.

01 / DECISION

Understand before acting

The problem
People need portfolio and market context before choosing a trading action.
My decision
I brought balances, market signals, and agent activity into one portfolio starting point.
Why it matters
A clearer orientation before moving into trading tools.
Portfolio dashboard
02 / DECISION

Make setup inspectable

The problem
Bot configuration can hide consequential choices inside unfamiliar settings.
My decision
I surfaced the exchange, strategy, risk selection, and investment amount during setup.
Why it matters
People can review the proposed parameters before activating a bot.
Configure an AI bot
03 / DECISION

Keep automation visible

The problem
After setup, users still need to know whether an agent is active and what it is doing.
My decision
I designed agent views to show state, runtime, trade count, and performance context.
Why it matters
Ongoing activity and controls are easier to find.
AI Agents

SUPPORTING ARTIFACTS

The structure behind the screens.

The project record documents MVP requirements review, competitor analysis, trader interviews, surveys, and personas. Original FigJam flows and a data model show how discovery informed account access, verification, recovery, and the product structure.

Original design artifacts · Figures inside the screens are sample interface data, not verified investment returns or project outcome metrics.

DESIGN ARTIFACT 01

Account access & recovery

Original FigJam flow showing PIN and password sign-in paths, notification checks, and recovery decisions.

DESIGN ARTIFACT 02

Sign-up & verification

Original FigJam flow showing registration, account review, verification, and the route into login.

DESIGN ARTIFACT 03

Data model

A map of NeuroTrade API entities, account relationships, and the operations that connect user, wallet, asset, and trading data.

VALIDATION & LEARNING

What I would validate next.

These are proposed next steps, not completed usability findings. No prototype-specific validation results, production launch, or financial performance outcomes are documented.

  • PROPOSED / 01Setup comprehension

    Can a participant explain the selected account, risk setting, and investment before creating a bot?

  • PROPOSED / 02Monitoring clarity

    Can they identify an active bot, distinguish realized from unrealized performance, and find its edit controls?

  • PROPOSED / 03Recovery & control

    Can they recover from a connection problem and understand the next available action?

What I delivered

Product UX direction, high-fidelity interfaces, core workflows, an interactive prototype, and supporting artifacts.

A complex financial product earns trust through clear context, inspectable settings, and visible user control.