Optimizing Efficiency: Stabilizing and Ongoing Technical Support for OFM Capital Partners’ System
Introduction
OFM Capital Partners is s a proprietary day trading company that trades US Equities, Options and Credit.
They have developed an internal system to collect data and provide valuable insight for the traders. Over time, this system became an integral part of their daily operations.
OFM Capital Partners were looking for a development partner, to support the operations of the system, while also adding additional features on top of it.
Being quite proactive and curious, HackSoft stepped in and became that partner.
The challenge
OFM Capital Partners' software is collecting a lot of data, from many sources. This data is then being processed and turned into important insights for the traders.
Over the years the software aggregated terabytes of data.
Additionally, the software itself was big and relatively complex.
And, as the cherry-on-top, the system was running entirely on-premise.
Being a mission-critical software for the daily business operations, we couldn’t just regularly take it down for a maintenance and upgrades.
So we had to draw from our software engineering experience and make a viable strategy for both maintenance and additional feature development.
The solution
We stepped in with a couple of experienced developers, full time, in order to review the current situation and make an action list.
We needed to setup the tooling for proper technical support.
Continuous integration
The first thing we did was to setup a proper continuous integration, running on GitHub Actions.
We started with the basics - run linters, formatters and deployment.
This ended up being a huge time saver because we were doing a couple of deploys a day for the next few months.
Error monitoring
Running a production system without proper error monitoring was like being blind. Something broke? We needed to be notified by the folks using the system.
That’s why setting up proper monitoring came high in the priorities.
Sentry was one of our go-to tools for that, which improved the quality of life by a lot.
Server Infrastructure
Initially, we thought of migrating the on-prem setup to the cloud, particularly - AWS.
After doing some quick math, we realized how expensive this move is going to be. The software was utilizing 16 cores and 64 RAM with around 2TB of disk space for the database.
That’s why we decided to continue running on-perm, while also purchasing additional hardware for redundancy.
Backups
Of course, daily backups were extremely important and we made sure everything was set properly, while also testing the recovery process.
Rewriting critical parts of the system
By having the logs and error monitoring in place, we quickly realized where the problematic areas of the software are.
First, we covered them with tests and after that, we proceeded re-implementing them in a more robust & stable way.
This approach yielded dividends:
The system started working smoothly, with no critical issues at hand.
We paved our way into adding new features to the system, without breaking everything else.
The result
Once the initial stabilization & feature development phase was done, we transitioned into a proper support model.
The system continued running smoothly, with occasional errors here and there, which we managed to address immediately.
HackSoft has proven to be a highly reliable software development partner.
Their engineering expertise significantly enhanced the overall reliability of our system while introducing crucial new features.
Beyond technical excellence, they demonstrated a strong understanding of our business needs - ruthlessly prioritizing tasks and ensuring that their work seamlessly integrated into our daily operations without disruption.
If you're looking for a development team that combines professionalism with peace of mind, I highly recommend HackSoft.
- Filip Madzharov, Managing Director
