Portfolio Sample
Situations and Events. It was in the aftermath of the 2011 Tsunami that hit Japan, that I worked at Caterpillar to help build the Assurance of Supply Center. It became apparent that such events can cause devastating disruptions to supply chains -especially when the SC is designed with little or no resiliency. These disrupting events can come in the form of natural disasters like earthquakes, hurricanes, floods, fires, but they also can include events like labor strikes at manufacturing locations and shipping ports, criminal, terrorist and geo-political acts, and as we've seen recently, in health pandemics like the 2020 Covid-19. In 2012, I worked with a team from Deloitte who created the playbook in what became known as "Situations and Events" to put into operation a way to monitor worldwide events and execute a quick response to any scenario. This project involved drawing together lots of data from both internal and external sources. We leveraged NC4 for monitoring events. We hooked up API's from public sources like National Oceanic and Atmospheric Administration (NOAA), the Joint Typhoon Warning Center (JTWC), the US Geological Survey (USGS), and others. We integrated this data with Google Maps and with Cat's geo-locations if its facilities and supplier locations. Then we built some data mining capabilities so that Cat could understand exactly what entities and transactions might be effected by the disruption. So for example, we could quickly map a hurricane to the exact parts on containerships on the ocean that may not arrive to supply a manufacturing schedule. We built a platform then to distribute this information to the supply chain people that were in a position to act.
Network Optimization. I managed another project for Cat in which we built network optimization models for its supply, production and distribution of certain machines. This was a forward-planning analysis that accounted for supply of parts, production capacities, and dealer demand, and asked and answered the question: Where should we make this machine? For example where Cat has demand for D8's all around the world, and is capable of making this machine in several places on several continents, where should it make the D8? Producing always where cost is lowest might seem right, but it has one big problem -time. Customers might not be willing to wait for delivery on a make-to-order machine. There are also a variety of constraints such as local content laws to consider. These are issues that we can't easily solve in a spreadsheet. So we build complex mathematical models that solve for the combination of production programs that meet the demand given all the constraints and at the least cost.
I managed a project for United Technologies (Now Ratheon Technologies) to design a network optimization model for their their worldwide production of wiring harnesses to supply the auto industry. This was an analysis of the company's production capacity (in North and South America, the Phillipines, etc.), and a modeling capability to set production according to demand of automobiles among UTA's customers.
Chaban Wellness, at the time was a small startup in the healthfood/diet and weight-loss industry. Chaban came to me needing to integrate their order-entry/billing with their distribution system. They needed a "middleware" solution so I developed and deployed an asp.NET application that processes order data, generates and sends shipping labels and packing slips to the fulfillment center. The solution had to do some analysis though. Customers were on subscriptions (through Zuora) and we had to evaluate the subscription and determine which items to include in the kitting operation. This proved to save our client a considerable amount of waste. Over time, I extended this basic system to validate addresses, and detect and manage suspected fraudulent orders. I supported Chaban in many promotional events, their transition to a different 3PL, different servers, and in their migration to a multilevel marketing model.
This HVAC, Sheetmetal/welding shop need some help processing their orders. I identified that the bottleneck in Therma's supply chain was in the process of preparing customer orders for fabrication. So I worked with a few key stakeholders to redesign the office processes. With some technology, we added automation by incorporating several 3rd party libraries and API's. We leveraged Office 365 and deployed an Outlook plugin. We tied this into a backend SQL Server to keep track of everything. In the end, we increased the throughput of the office bottleneck by over 200%!
For a period of about 10 years, I managed a number of projects for Boise Cascade, to design, develop, deploy, and support Supply Chain software systems related to the manufacture and distribution of wood products. These were Business Process Re-engineering/Business Transformation projects that included 1) General Optimization Modeling system for purchasing/production/marketing of all lumber products and of all facilities, 2) Master Production Scheduling systems for Engineered Wood Products Manufacturing facilities, 3) Development of BC TRACKER™, inventory software distributed throughout North America. These systems used the following technologies: VB 6.0, Javascript, MS Access, MS SQL Server 2000, XA Linear and Mixed Integer Optimization Solver, MS MapPoint GIS.
I designed, developed and implemented an optimization modeling system for the Perdue Farms turkey business; This was used to analyze and plan the marketing of commodity, retail and foodservice turkey, and further processed turkey products; I generated a linear programming model to describe the availability of the meat supply, expected parts yields, product bill-of-materials, processing costs and demand quantity and price for every manufactured product at every facility. The solution gave us much insight into the economics of the business.