Click outside the image or press ESC to close
We design and develop scalable software solutions, backend platforms and automation systems for modern businesses.
Balani Software Services (PTY) specializes in custom software development, backend systems, API integrations, algorithmic trading systems and business automation.
We focus on delivering maintainable, scalable and performance-driven solutions tailored to client requirements.
We are a CIPC and SARS compliant fully black-owned business who bridge the gaps between high quality delivery and competitively priced services.
'I am not a Ghandi nor a Mandela, I am in the business of selling software to people who recognize value and are happy to purchase it.
I am not interested in building software for the price of a coffe cup.'
- Monwabisi Apollo Balani
REST APIs, business logic, authentication systems, microservices and data processing.
This is our speciality. We build every single type / kind of indicator, trading tool, automated system, risk model, data capture, data sourcing etc.
Almost every platform is supported (TradingView, MetaTrader, NinjaTrader, ThinkSwo, QuantTower, Sierra and many more)
Internal tools, workflow automation, reporting systems and integrations.
With a large focus on building both client facing and custom admin applications for internal tools.
After extensive experience in building software for clients we offer consulting for clients who may not be certain what architecture or strategies they want implemented.
Disclaimer: Our consulting services assumes no proprietory information is shared by the client unless explicitly specified.
therefore, in cases where non-copyright assets are re-used, no consent will be required from a previous client.
Problem: Client needed a full overhall of an HFT system to achieve faster trade execution speeds to
capture more edge and less slippage under high volatility market conditions.
Solution: Wrote DLLs using C++ to perform algorithmic calculations and pre-execution data to allow sub 3ms server side trade execution. The main bottleneck just happened to be MetaTrader 4 cannot conduct use multi-threading, this solution circumvents that issue by performing the calculations simultaneously outside the MT4 terminal.
Outcome: Achieved sub 3ms trade execution, which is a significant change from 40ms as of previously. Developed a low latency tick-only charting algorithm for client side visuals (as seen in screenshot)
2026.06.11
Problem: Client needed a user friendly application that sends alerts via email and a GUI that has graphing elements that can fluidly display a multitude of data points and metrics.
Solution: Wrote a comprehensive application that uses primarily matlib to reliably display all the specified metrics and data.
Made use of Qt library and matplotlib to chart data, and FastAPI integration to fetch server side account data to retrive values such as Running PNLs, open trade postions, trade history etc.
Outcome: Achieved real time updating chart functions in the external application to help the client maintain up to 100 trade accounts using a heat map. Also the program can send real time custom alerts / notifications directly to email using SMTP.
2026.04.21
A demo copy is available for most of our paid pre-built models / software.
For auto traders and Indicators, the trial only runs in testing enviroments through an a time-based restriction, usually the first of current month.
To enable maximum freedom to test our software outside the Strategy enviroment all our software can be applied to chart but trading cannot be enabled on trial copies.
Strategy: This is a rudementary psuedo HFT system. The underlying algorithm identifies the
average fair value price of the Volatility 75 Deriv.com synthetic index.
Synopsis: The initial idea behind this strategy was an attempt to reverse-engineer the price of the instrument. Because it is synthetic, it is technically possible to reverse engineer the price. I just so happened to not quite make it that far, but the results say the system is close enough to be absurdly profitable (during low spread / high volaitility periods)
Features & Inputs: [See image 2 for screenshot]
Tester Results: Running the EA in a live account has the disadvantage of being significantly more sensitive to slippage and spread variances. So, strategy tester results are not to be taken as reprisantive of real trading enviroments. This is not to say the EA is unprofitable, obviously, otherwise it wouldn't be featured here. Just to say a 99%+ Z-Score / a 7+ Sharpe Ratio is obviously vastly innacurate, which is how this tests. [See images for screenshots]
Strategy: This is a starightfoward Auto Trader that is based on the (disputed) concept of the so called 'market fair value gap', popular amongs 'SMC' and ICT traders. The short of it is that the market continuously creates price imbalances and those can be spotted by large expansions away from a price level. Making them key entry levels for later on.
Synopsis: Why I, the dev, decided to build the EA was mainly to see if
Features & Inputs: Basic Trade Management inputs. Use designated default magic number to run EA smoothly and avoid order collisions. [See image 3 for screenshot]
Tester Results: I would never run such an EA in a live account simply because the fundimental strategy is incorrect even though correct in theory.
YES: price does create imbalances and later retracts to correct for them, and yes a reliable model can be built around that.
NO: FVGs and OrderBlocks, or really any ICT/SMC adjacent system / model has absolutely no predictive power in determining where those levels are.
Funny enough, this specific EA, under the right conditions does go on decent winning streaks. This alone however is not significant to offset the losing streaks or the fact that it does not beat the baseline strategy.
[Think coin-toss models; example: an EA that has 100 point TP, 100 point SL. One trade per day. -> using an RSI + ATR + ADX combination to determine the most optimal price action and taking an order, the EA works out to about 60/40 WLR. In a live account that would average down to about 50/50 after slippage, fees and spread.] Any system that cannot beat that minimum, is considered unprofitable, and this fits that criteria. However I am not in the church / religion business, hence I do not tell people what to believe. Hence I have made this EA available for free.