60 different ML techniques were used to predict demand of SKUs as per past data. Automated EDA used to format raw data as per requirements.

I did the project earlier in 2016. In 2018, my client called me again to implement Benders Decomposition, Column Generation etc to reduce the computational time of scheduling problem for about 5000 call center agents,

• It is a capcitated vehicle routing problem with due consideration on time window, draft, and other drastic parameters.

Technique: Mixed Integer Programming, Heuristic Method, and Data Clustering. Data Clustering was used to cluster route of each loading port by using shortest neighbor distance and then sub-route algorithm was applied accordingly.

Tool: CPLEX & Python and CPLEX CONCERT with Java/C#

Language: Python and R