Data Mining – Moving Towards Actionable Output with Automation

Data Mining has paved a way for answering many questions which were unanswered in the past. Data Mining helped to do a lot of root cause analysis and helped industries to improve their productivity by understanding non-value add tasks and waste reduction etc. In addition, helped to improve sales by looking at customer buying pattern and market analysis.

This has been very exciting and helped business grow and created opportunities to improve.

In addition to this it also raised questions on how we can improve the current situation for business with data mining rather than looking backwards and/or forecasting future planning(which have lower credibility due to COVID situation) with AI and ML.

Can we make life simpler/easier for business with Data Mining and with possibilities of Automating the processes?

In one of my previous experience, we had a thought that how we can prioritize thinks for business e.g.

• which orders to release first?

• which customers to serve?

• which products to order first?

• if we have blocks due to many reasons, which one I act on first to keep the delivery dates?

These questions are answered in data and is generally available in your system and business use that to make those decisions.

Question was, why we must do this manually when we already have the data to prioritize things and best why cannot we automate that.

These questions made us think how we can leverage combination of Data Mining, AI, and ML with Automation to make things efficient and effective for business and ROI for Data Mining.

Data Mining use case was showing ROI as below:

At that point of time, I came with an approach that we could switch to different ROI Model.

With this concept, we used Data Mining to not only look backward or forecasting but look at current business, profile orders and blocks and credit management using AI and ML to identify current bottlenecks in business and de-bottle neck them.

What we provide to business is an operational view of

• what is happening currently in their part of responsibility,

• help them to prioritize the work,

• automate some of the non-value add work,

• link different knowledge and information point from ERP, CRM, and satellite system to provide them a view which becomes their starting page in the morning and day end closing view.

An example would be to provide business:

• how many orders need to have deliveries created today so we can meet the requirement for the customer delivery date.

• Which of them are in any block like, credit block, delivery block etc.

• Triggering emails to responsible people or if there are rule set like removing credit block based on certain criteria defined by FIN colleague (SMART credit management) – automating that part of the process.

• Use Data Mining to look for delivery options based on routes, timeframes of delivery and find out when we should start our delivery process and for which orders.

Data Mining, AI and ML together with Automation helped business to gain that insight and make it actionable right away. Which in addition to the root cause analysis provides rich dividends and keep the business engaged.

Conclusion

What we learnt here that keeping engage business and management for long running data mining projects which provide great insights Is not enough. We need to find a solution which can help to make life easier for business on operational level.

My advice to Data and Analytics leader would be to look for synergies with local business units on what they are doing and make them your allies. Enable and empower them so they can contribute to not only local success but also to global success of the organization.