Contact us now: Phone: +27799778342




Business problems are probably the only constant in a business. Problems seem to be everywhere and no company is excluded. Finding solutions to these problems seem like daily activity for most people, however at Autolytix Data Science we approach problems in a new and innovative way.

Problems are just constraints in your business. The fundamental principal of solving problems are to eliminate the constraints. New constraints will soon follow, but removing the constraints in your business will allow you to grow. The Theory of Constraints, was developed by Dr. Eliyahu M. Goldratt. He was an Israeli physicist who changed the way people think about their businesses. Click here to learn more about Theory of Constraints.

Similar to Dr. Goldratt, we have developed our own framework to solve a problem permanently. Autolytix Data Science is results focused and because of this we are able to clearly define a plan and the execution that follows.

How we solve business problems at Autolytix Data Science


Think about the result you want to achieve and set your goal or objective. Be specific and quantify the result, i.e. $200m in revenue in the next quarter.
Here you identify the problems or constraints standing in your way of achieving your goal. Be specific and quantify the problem, i.e. lost sales has has cost our company $30m in the last quarter. If we reduce lost sales by 10% we will increase revenue by $3m.
Problems can get bigger over time. That's why you need to think about it now. What happens when this problem scales and becomes bigger than we anticipate? Again, quantify the the scaled problem. If lost sales doubles we will loose $60m in revenue over the the next quarter.
Identify the underlying variables causing the problem to exist. Normally in this phase you will do some research and analysis to quantify the underlying variables.

For example you'll find that there is high volatility in incoming sales orders. Some days you have to many orders to fill and other days are quite and you have excess capacity. Or maybe you are producing the wrong product and at the wrong time. Or maybe a combination of both.

Quantify the role of each variable and to what extent will it influence the scaled problem,i.e. random orders account for 80% of the contribution to lost sales and incorrect production accounts for the remaining 20%.

Assess the push and pull levers affecting these variables and to what extent you can influence them. The solution to sales order volatility is either to smooth the sales orders or adapt to the volatility. Demand planning, that informs production planning will solve the problem with production. Both problem can be addressed with the correct analytical software. Firstly an intelligent sales order application can be developed to take client orders. It should assess the order size against available capacity and correctly schedule a date when the order can be 100% fulfilled.

A production planning application can be developed to intelligently look at client sales orders and lost sales to sequence and plan future production so that it provides the correct inventory cover to satisfy demand.

Develop a project plan with the required resources to develop, test and implement the plan.Here you also need to do a project costing. The cost-benefit analysis is essential before you proceed with the project. One important thing to remember is to keep your project team as lean as possible. The fewer people involved, the higher the probability for success.
Start developing the solution. Start by addressing the biggest problems first. The solution can be anything. It can be software, a business process, company policy, training or an Excel spreadsheet. In our example this is where you will model and develop the algorithms, develop the back end, including the database as well as the front end application.
Here you'll need to affect change. People will directly affect the success of the implementation. The best way to affect change is to be inspired. Inspired people will embrace the change dilligently execute the project.
Measure your performance against the original goal. Did we achieve it? Where did we fail? What do we need to change? Go back to step 1 if you where unsuccessful.
This is an essential part to solve a scaled problem. You do not want to address this problem again in the future, and the only way to do it is to automate the processes for self governance. You only need or want to know about it when exceptions to the rule occur. Data Science is the perfect medium for automated self governance, for example, an algorithm can be written to send out an email to certain people when the business rules are breached. In all other cases you will only receive a daily email with the summary results. All automated.
When the entire solution is automated you can move on to a new challenge. Again start with step 1.


Following these steps will allow you solve business problems in a scalable fashion, that governs itself, so you can focus on other areas of the business.