← All selected work

Risk-Controlled AI Trading Agent

Turns financial news into paper-trading decisions with human approval

Paper-trading prototype · 2026
Local paper-trading dashboard showing live signals, pending approvals, news activity and system health with account amounts hiddenInspect screenshot ↗

The problem

A news signal is not a trading decision. The project explores how to connect language-model analysis to technical checks, explicit risk limits, and a person who can inspect the reasoning before approving an action.

My contribution

Personal project: news ingestion, signal analysis, approval workflow, risk controls, and a review dashboard.

  • Python
  • Redis
  • Technical indicators
  • Paper-broker integration

What I built

The approach

Reads financial news, confirms the signal with technical indicators, and prepares a paper trade for human approval. Every decision is logged for review.

Read the source

Operating settings & risk controls

15 minnews cycle
5 minposition checks
10%single-trade cap

These numbers describe configured behavior, not investment performance. Trades require human approval; the repository does not establish profitable returns.

Follow the workflow

News is collected and queued for analysis.

Inside the project

Saved screenshots from the repository. Select an image to inspect it in detail.

Engineering decisions

01

Keep a person in control

A person must approve every trade before it reaches the paper account.

02

Do not trust news alone

Signals must also clear technical checks, reaching 55% combined confidence.

03

Limit damage before chasing returns

Trades are capped at 10% of equity, exposure at 50%, with a $400 kill switch.

Limitations & next steps

Add repeatable backtests and a paper-trading report before making a claim about signal quality or returns. Evaluate rejected signals as well as approved trades to understand what the controls actually change.