Supply Chain Optimization

Inbound warehousing process and inventory optimization using Kobai's semantic layer integrated with Databricks.

Helping CPG (Consumer Packaged Goods) and FMCG companies with supply chain planning and execution. Providing an operations planning digital twin and an operations execution control tower for impact analysis of decisions on specific KPIs.

About the Client

A leading multinational corporation that specializes in the production and distribution of a diverse range of consumer food products. With a global footprint spanning multiple continents, they operate state-of-the-art warehouse and retail facilities equipped with the latest automation and robotics technologies.

The Challenge

  • Prioritization Complexity: The process lacked a dynamic prioritization system for incoming shipments. Factors like POs, labor schedules, capacity, demand, promotions, and allocations were not fully integrated leading to manual intervention.
  • Limited Data Integration: Existing reports relied on static filters and did not capture the real-time dynamics of warehouse operations. Lack of comprehensive data integration hindered informed decision-making regarding prioritization and resource allocation.‍
  • Capacity Constraints: Incoming shipments often exceeded available warehouse capacity, creating bottlenecks and delays. The system did not provide clear visibility into workload and capacity limitations and affected resource management.
  • Manual Workarounds: The reliance on manual filtering and prioritization in reports created potential inconsistencies. This caused delays, inaccuracies, and difficulty in monitoring the overall process effectively.

Kobai's Approach

Kobai’s collaboration with the corporation aimed to implement an end-to-end supply chain planning solution which includes an operations control tower and operations planning, leveraging the power of semantic graph layer integrated with Databricks lakehouse.

The team worked closely with the data team to enable the quick integration of diverse data sources like PO, Inventory, Labor availability, Demand, and Promotions into a single, flexible semantic model, in a collaborative, self-service way. Data complexity within the context of PO receiving and its complex relationships were analyzed, thereby identifying patterns and failure modes, leading to improved PO receiving, labor planning and product priorities.

As a part of our approach, customized ontologies and semantic graph tailored to the corporation's supply chain domain, capturing domain-specific relationships and constraints were developed. This aided Smart Collaboration, Decision Support, Decision Scenario Planning, and Autonomous Action.

Moreover, Databricks' DIP enabled the execution of complex analytics workflows, including demand forecasting, route optimization, supplier performance analysis, and inventory optimization algorithms. Not just that, Databricks notebooks and visualization libraries were used to create interactive dashboards and reports, providing stakeholders with intuitive insights and actionable recommendations.

The Impact


The implemented semantic graph delivered significant benefits:
  • Seamless execution of operations and improved decision-making through real-time visibility into supply chain operations.
  • Streamlined reactive and predictive planning and smart alerts.
  • Enhanced agility to respond to market fluctuations and demand changes.
  • Reduced inventory costs and stockouts by optimizing inventory levels and replenishment strategies.
  • Increased efficiency in logistics and transportation management and proactive risk management through predictive analytics and scenario planning.

By leveraging Kobai's semantic graph integrated with Databricks, the corporation successfully optimized its supply chain operations, achieved optimum market management and cost savings, and enhanced overall efficiency.