From trading engine to live copy trading platform

From trading bots to a customer-facing service

A team of traders and analysts approached me to build a copy trading platform. They already had their own trading bots and wanted to offer them to other users, growing revenue by taking a share of customers' trading income.

Users would register, connect their exchange account through an API, and choose a bot and an acceptable risk level. The bot would trade in their account, and they would pay a percentage of the income it generated. A free trial let them assess the bot's trading in their own account before deciding whether to pay to continue.

The client's team, primarily one of the founders, developed the trading engine. My job was to turn it into a customer-facing product with a portal, bot connections, user account management, data collection, and reliability mechanisms.

Connecting a bot and renewing access

Customers arrived through advertising, registered, connected their cryptocurrency exchange account through an API, and selected a trading bot. The bot made trades during the free trial, so users could see the results directly in their exchange account.

If they wanted to continue after the trial, they paid to renew access. Periodic payments kept access active; if the next payment was not made, the bot was disconnected.

Users could choose one of four exchanges. During setup, a slider let them adjust the balance between acceptable risk and desired return. This setting controlled trading parameters, not a promised level of income.

That straightforward customer journey required several mechanisms to work together: account and trial tracking, payment processing, access renewals, and bot disconnection. I connected these into a complete customer service rather than just an interface to a trading algorithm.

My work on the platform and product

I started with the architecture and customer portal prototypes, then coordinated a designer who created the visual layouts. I handled the frontend implementation, backend programming, and debugging myself, and continued developing the platform after handing it over to the team. I also organized data collection.

Beyond implementing features, I contributed to product decisions. My experience with websites and interfaces helped me identify usability issues and suggest changes. The client adopted many of my proposed improvements and optimizations.

In a separate assignment, I researched trading strategies, including arbitrage, and prepared reports on my findings.

Reliability when market data feeds were disrupted

The main technical challenge was external market data. Exchange feeds were not always reliable, yet the bots trading in users' accounts depended on them. For a service paid from trading results, those interruptions posed a significant risk.

I worked on reliability from the start, addressing service availability and stability requirements and adding redundancy to the retrieval of exchange price data.

Later, I also helped improve the trading engine's resilience. Working with the founder, I contributed to redundancy, intelligent caching, and other performance and stability mechanisms. The original trading engine remained the client's work.

Launching with real users

The platform launched, and the team began attracting customers. Real users connected bots and traded through them in their own exchange accounts. The service operated for more than six months.

If you need to turn an existing algorithm into a customer service, or build a portal with subscriptions, free trials, and payments, see my internal systems and web services offering.