Worked Aurora deployment guide

Deploy an AI trading bot to Public

Build a GOOG RSI rebound bot with Aurora, inspect the fixed-rule backtests and add a separate recurring research review before moving the trading rules to your broker.

As of 2026-10-10

Follow one worked path

validate candidate workflow diagram with review boundaries
Source-derived workflow diagram, not a screenshot or an external-client execution receipt.
Step 1 of 4

Review this task's exact candidate

Open the prepared Aurora request. Review the candidate rules below, including its entry, position guard, purchase amount and exit. Continue to a saved research draft only after reviewing those rules.

What you should see

Aurora shows the candidate rules and returns the saved research draft identity after you approve creation. The brokerage connection comes later.

backtest result workflow diagram with review boundaries
Source-derived workflow diagram, not a screenshot or an external-client execution receipt.
Step 2 of 4

Read historical and paper evidence separately

Follow creation, historical testing and paper observation in the same conversation. Retain the returned saved draft, test and running paper identities.

2023-03-30
Test ends (Eastern): 2023-12-06 · Return: 2.85% · Max drawdown: 5.00% · Closed trades: 10
2023-12-07
Test ends (Eastern): 2024-08-14 · Return: -0.36% · Max drawdown: 8.77% · Closed trades: 10
2024-08-15
Test ends (Eastern): 2025-04-23 · Return: -1.61% · Max drawdown: 13.33% · Closed trades: 10
2025-04-24
Test ends (Eastern): 2025-12-31 · Return: 19.14% · Max drawdown: 2.83% · Closed trades: 12
What you should see

Each stage has a separate artifact to inspect. A simulation curve is not broker account activity.

broker review workflow diagram with review boundaries
Source-derived workflow diagram, not a screenshot or an external-client execution receipt.
Step 3 of 4

Connect and review the broker target

Connect Public using the account steps below. From the tested Paper portfolio, choose Port to Live and review the added and disabled rules before confirming.

Open Public connection
What you should see

Confirm & Port can activate the live target. Inspect its account identity, trading policy, orders and active state.

broker review workflow diagram with review boundaries
Source-derived workflow diagram, not a screenshot or an external-client execution receipt.
Step 4 of 4

Verify execution and the pause control

Match portfolio events with broker activity. To pause strategy evaluation switch Automated Trading off and Save Changes in the live Deployment Center.

What you should see

Outstanding broker orders and holdings require their own review; pausing does not automatically sell or cancel them.

Create and test this example in Aurora

Show create and test this example in aurora

Open the filled-in request, review the rules and send it in Aurora when ready. Continue in the same conversation through creation, backtesting and paper observation. Each step returns an artifact you can inspect before continuing.

Understand the worked rules

Show understand the worked rules

Scroll sideways to compare all columns. Select a heading to sort.

EnterGOOG 14-day RSI < 30 and no GOOG holding.
SizeBuy whole shares with 100% of available buying power.
ExitGOOG RSI14 > 50 with a positive holding; sell the whole position.
AI's jobAurora explains, creates and reviews the saved rules. The trading engine evaluates those rules.
Optional recurring reviewA separate Friday research agent reads sources and proposes revisions. It leaves the trading rules unchanged until you review a new version.

Read the completed historical example

Show read the completed historical example

These are actual fixed-rule GOOG RSI rebound backtests. Aurora's future research suggestions are outside this historical test. Compare the completed rules and trades before you decide whether to keep this example or create another version.

Scroll sideways to compare all columns. Select a heading to sort.

2023-03-302023-12-062.85%5.00%10
2023-12-072024-08-14-0.36%8.77%10
2024-08-152025-04-23-1.61%13.33%10
2025-04-242025-12-3119.14%2.83%12

Observe it in paper trading

Show observe it in paper trading

After testing your frozen version, ask Aurora to prepare a new paper deployment from the saved draft. Read back the new simulated portfolio identity and open it in Portfolios. Log the expected signal, actual order status, holdings, fees and any rule revision. Keep the historical backtest, paper history and live account history labeled separately.

Add a separate weekly Aurora review

Show add a separate weekly aurora review

Open this request after choosing the bot's saved strategy. Review the Friday schedule and research scope in Agents before enabling it. An enabled recurring review consumes research credits when it runs; its history records each run. You can pause it independently of the trading portfolio.

Connect and select your Public account

Show connect and select your public account

Open Public account connection below. Select Connect Public, then Connect my existing account. Complete two-factor authentication if prompted. In Important Disclosures, select I understand and agree, then Connect Public and finish authorization on the broker's site. Return to NexusTrade and let Checking connection finish.

Complete Public's account requirements shown in Connection health, then compare the imported account, buying power and holdings with Public. Select the live Public portfolio as the target for the saved paper strategy.

This example trades whole stock shares. The current NexusTrade adapter also supports fractional stock shares. Keep the chosen account ID with your deployment notes.

Move the paper rules into the live portfolio

Show move the paper rules into the live portfolio
  1. Open the paper portfolio

    Use the separate paper portfolio created from your saved draft. Open its Deployment Center and select Port to Live.

  2. Select Public and inspect the changes

    Read the incoming rules and existing active strategies. The confirmation shows which existing rules will be disabled. Keep the paper source so you can compare later observations.

  3. Confirm the intended deployment

    Select Confirm & Port when the comparison is correct. If active strategies already exist, the button reads Disable existing & Port. This confirmation can activate the live target.

  4. Open the returned live portfolio

    Select Open live portfolio. Verify Public, the intended account and the imported rules. Review the live target's evaluation mode after porting. OpenClose checks near the session's open and close; Constant checks repeatedly at the chosen interval. Keep the observation interval and evaluation cadence in your test notes. Port to Live copies strategies; it does not copy the paper portfolio's timing settings.

Follow the first signal and know how to pause

Show follow the first signal and know how to pause

Review the live portfolio's trading policy and daily limit in the app. Only you can enable automated order approval through the account application or portfolio settings. Aurora can help inspect the rules and events; it does not enable that policy.

Follow the RSI entry, Filled buy, later RSI exit and Filled sell in the live portfolio. The recurring research agent has its own schedule and history: read its sources and proposed changes separately from the bot's trades. Accepting a research suggestion creates a new strategy version to test.

To pause strategy evaluation, open Deployment Center, switch Automated Trading off and choose Save Changes. Check that the same live portfolio is paused. Inspect open broker orders and current holdings separately. For a recurring Aurora review, also pause its schedule in Agents.

Continue exploring

Discussion

Sign in or create a free account to join the discussion.

No comments yet.

Aurora · AI research assistant

Try this in Aurora

Edit the request, then open it in Aurora. You choose when to send it.