Products Observe

See every fill as it happens.

A self-hosted observability dashboard for qkt. Every trade, order, log line and equity curve, from every instance, in one place.

View on GitHub

Licence: Apache-2.0
Needs: qkt 0.40 or newer
Storage: SQLite, self-hosted
Image: ghcr.io, public

Replay of a backtest · BTCUSD 15m · August 2026

One trade, in order. Click an event to see it on the curve.

Trade 1 −63.04 USD

Trade 2 +125.62 USD

Trade 3 +127.51 USD

Trade 4 −64.97 USD

Trade 5 −64.04 USD

Trade 6 −63.62 USD

Trade 7 −63.37 USD

Trade 8 +125.84 USD

Trade 9 +128.61 USD

Trade 10 +138.67 USD

Trade 11 −77.47 USD

Trade 12 −77.30 USD

Trade 13 +153.11 USD

Trade 14 +157.99 USD

Trade 15 −79.52 USD

Trade 16 −78.79 USD

Trade 17 +156.91 USD

Trade 18 −79.94 USD

Trade 19 +155.67 USD

Trade 20 open at the end

In motion

A trade, as it happens.

Replay one trade from a backtest: the rule fires, the order fills, a bracket is set, and the equity follows.

The first trade of a backtest replayed: the rule fires, the order is sent and filled, a bracket is set, the stop hits and equity follows.

The real dashboard

Every fill, as it lands.

A real screenshot: a paper session on Deribit's live prices, run with qkt 0.56.0 on 12 October 2026. Two round trips at 0.001 BTC, one stopped out and one at target. No real money was involved.

The qkt Insights Trades view for instance demo-paper: four fills of the fast_cross strategy on PAPER:BTC_USDC_PERPETUAL at 0.001, a stop exit at minus 0.06 and a target exit at plus 0.04.

Out of the way

Observability never touches the trading path.

qkt pushes its events to Insights over HTTP, from its own thread. If the dashboard is down, trading carries on, and events are kept on disk until it returns.

Account and equityBalance, equity, open positions, drawdown, month by month.
Every order and fillSearch and filter the deals, with the strategy that placed each one.
EdgeWhen a strategy makes money, by weekday and hour, greyed out where there is too little data.
Health and logsLive log tail, uptime, and a record of incidents.

Run it

One container.

Point qkt's insights config at it with the ingest token, and the dashboard fills in.

run insights
docker run -d -p 127.0.0.1:8420:8420 -v insights-data:/data \  -e INSIGHTS_DB=/data/insights.db -e INGEST_TOKEN=... \  -e ADMIN_USERNAME=admin -e ADMIN_PASSWORD=... \  ghcr.io/elitekaycy/qkt-insights:latest run

Share a result

Public links that keep your account private.

Share a strategy or the whole overview with a link. Shared figures run 15 minutes behind, no cookies are set, and visitors are counted once per day without storing IP addresses.

Read about sharing

The path

1Write and testqkt engine
2Connect a broker or venueMT5 gateway Venue gateway
3Set limits on the accountGuardrails
4Watch it runqkt Insights

Ready? Follow the step-by-step guide

Write your first strategy today.

One command to install. Eight lines to a backtest. Free and open source.

Get started Read the docs