Bring-your-own-algo command center

You built the algo.
This is its cockpit.

QuantumEdge is a self-hosted command center for any trading algorithm — live, paper, and backtest side by side, on your own IBKR infrastructure. Your strategy stays yours. The instruments come from us.

3 modes · one screen ~3 min self-heal 100% self-hosted 0 strategy code shared
IBKR Connected
Portfolio · your algo · live
$52,148
+$318 today
▲ click the modes — same dashboard, three realities
IBKR-native · Gateway 4001/4002 Self-hosted · your server, your keys Admin + guest roles · investor view Watchdogs · self-healing sessions

One strategy. Three realities. One chart.

Every algo lives three lives: the backtest that made you believe, the paper account that made you careful, and the live account that pays the bills. Most traders keep them in three different tools and never see the moment the stories diverge. Here, divergence has nowhere to hide.

Cumulative return · first 16 weeks of each mode
— sample data · hover the chart · click legend to isolate · your dashboard renders your numbers

Backtest Simulated

What the strategy promised — with every assumption it made baked in. Load your results file and it renders in the same charts as everything else.

Paper Forward test

What it does with real prices and real order mechanics — the honest rehearsal before real money, tracked continuously, not in a spreadsheet.

Live Real capital

What actually happened. The only line that compounds — and the one every other line exists to protect.

Backtests tell stories. Comparison tells the truth.

Walk the deck.

Every panel below is a live mock of the real interface — click through the tabs. Built by algo traders who got tired of tailing log files over SSH; every screen exists because we needed it in production.

Plug in your algo. Keep your edge.

No rewrites, no framework, no sending your strategy anywhere. If your bot can write a file or reach an IBKR Gateway, it can drive the full dashboard.

1

Connect

Point the dashboard at your IBKR Gateway (paper or live), or have your bot drop simple JSON files — or both.

2

Configure

Name your algo, set its schedule and rules metadata. The interface takes its identity — this is your cockpit, not ours.

3

Command

Monitor positions and P&L in real time, review every signal, pause or kill from anywhere, and invite investors as read-only guests.

A small, documented file contract — language-agnostic. Python, C#, Pine-exported, hand-rolled: if it trades, it plugs in. Any file you don't provide, that tab simply waits.

The upgrade is the whole workflow.

Running blind

  • SSH in and tail -f bot.log to learn what your money did
  • Backtest in a spreadsheet, live results in a broker app — never on one axis
  • Gateway dies at 11:45 PM; you find out Monday
  • "Why didn't it buy today?" is an hour of log archaeology
  • Investor wants to see performance → you screenshot your terminal

Running QuantumEdge

  • Positions, P&L, and equity history on any device, live
  • Backtest, paper, and live in the same charts — divergence visible day one
  • Watchdogs restart a dead session in ~3 minutes, logged
  • Signals tab shows every candidate and why it was taken or skipped
  • Investor gets a read-only guest login in 10 seconds

Built to survive the night.

IBKR kills sessions nightly by design. Most home-rolled setups die with them. Ours doesn't — this layer alone is worth the switch.

Layer 01

Nightly re-authentication

The Gateway restarts itself on schedule and logs back in with stored credentials — IBKR's forced daily cycle happens on our terms, not at random.

Layer 02

Port watchdogs

Independent watchdogs probe the Gateway, API, and bot every few minutes. A dead session self-heals in about 3 minutes — whether you're watching or asleep.

Layer 03

View-only live wiring

Runs on your server with your keys, and live accounts can be wired read-only at the IBKR level — the dashboard physically cannot trade what you don't allow.

Questions traders actually ask.

Does my strategy code ever leave my server?
No. The dashboard reads your algo's outputs — JSON files it writes, or account state from your own IBKR Gateway. Your source code, your parameters, your edge: never uploaded, never requested, never seen.
What do I need to run it?
A Linux server or VPS (a $12/mo droplet is plenty), an IBKR account, and IB Gateway. We ship the dashboard, the API layer, the watchdog stack, and the setup runbook. If your bot already runs on a server, you're most of the way there.
What does my bot have to output?
As little as one small JSON file. Signals, positions, trade history, and backtest results each have a tiny documented schema — provide what you have, and the matching tabs light up. Any language that can write a file qualifies.
Can the dashboard place trades on my live account?
Only if you allow it. The recommended live wiring uses IBKR's read-only API mode, which makes order placement impossible at the broker level — the dashboard becomes a pure instrument panel. Paper accounts can keep full bot control (pause, resume, kill switch).
Which brokers are supported?
Interactive Brokers natively (paper and live, via Gateway). The file-based mode is broker-agnostic — if your bot trades elsewhere and writes JSON, the portfolio, signals, trade log, and backtest views all work.
Who is this for?
Solo algo traders and small teams who have a working strategy but no instruments around it — and anyone who needs to show performance to a partner or investor without handing over broker credentials.

Your algo already works.
Give it a command center.

Access is granted per deployment while we onboard early operators. Tell us what your algo trades and how it runs — we'll take it from there.