Business Objective: Drive efficiencies in the Supply Chain for a large manufacturing company and recommend optimal spares stocking at central repository, branches and service franchisees, aiming to reduce holding cost without impacting service levels. Key Solution:
- The model forecasts spares consumption across the install base (existing as well as expected)
- An ensemble approach is used for the Predictive Modelling (using a library of algorithms) that auto selects the best fit model for each spare
- The model factors in consumption trend variation by season, geography and machine type
- Based on the expected consumption and supply side parameters such as inventory holding capacity, cost constraints, lead times, etc., the optimization engine recommends the optimal spares stocking at various points in the Supply chain using a Mixed Integer Programming approach
Levers from tcg mcube:
- High volume data ingest capabilities from a variety of sources
- Elastic for In-Memory to gain rapid search and query response
- Spark for distributed/ parallel processing
- SparkR algorithms for AI & Machine Learning
- tcg mcube Visualizations
Business Impacts & Outcomes:
- Helps in operational planning and minimizes the risk of stock outs, balancing inventory holding costs
- Increases customer satisfaction through reduction in call closure delays due to insufficient spares availability at franchisees
