Trading terminal and trading software development
Internal platforms that let a small team run many trading bots and change strategies safely.
A trading terminal for operations is the screen and the rules behind dozens of bots. We build trading software in Python around a strategy language the operator edits, checks that reject a bad parameter before it goes live, a live PnL feed and paging that reaches the person on call.
When to build trading software
-
Many bots, few people
Strategies change often, each server has its own setup, and one wrong parameter can lose money fast.
-
One strategy on one account
A broker's own platform or a single script is enough at that size.
What the terminal includes
-
Strategy language
A Python DSL in which the operator writes and edits strategies.
-
Checks at admission
Parameters are validated when a strategy is admitted, before it trades.
-
Pattern mapper
Chart patterns are linked to the strategies that act on them.
-
Trading dashboard
Live PnL streams over WebSocket, with the working state in Redis.
-
On-call paging
Telegram pages the person on call with the PnL attached.
-
Daily deploys
A CI pipeline releases every day, and each release passes a suite of strategy checks.
Case study: one operator, dozens of bots
-
Where the risk sat
Parameter files and deploy steps differed from bot to bot. A strategy change needed an engineer, and live PnL sat in a spreadsheet.
-
What the platform changed
Strategies live in one Python language and are checked on admission. PnL streams live, and Telegram pages the person on call.
Built with Python, WebSocket, Redis. Read the case study: Trading operations terminal
Trading software we have built
-
Trading operations terminalOn the platform we built, one person looks after dozens of bots across dozens of servers and ships strategy edits without an engineer.
How a trading software project starts
-
Bots and servers
Describe the bots you run, where they run and what a mistake costs.
-
Rules and checks
We agree what the operator may change and which checks guard each change.
-
One strategy live
A first strategy moves onto the platform with its checks, and the rest follow.
And much more. Ask about your project
Common questions
What is a trading terminal?
Can an operator change a strategy without a developer?
Do you build algorithmic trading software or trading bots?
How many bots does one person run?
Tell us how strategies change today and who gets paged when one goes wrong.