Request Expert

RETAIL FORECASTING; Expert D. Expert has 25 years of experience developing forecasting techniques that support specialized optimization systems for apparel manufacturing, including material utilization, inventory management, retailers' orders, and store-level and aggregate retail sales. These problems involve non-symmetric processes, "noisy" store-level demand, dominant customer, and results-to-date analysis (similar to election-night analysis). The approach uses multiple sources of information and analyzes details and fluctuations rather than just using totals or averages; it emphasizes clean input data plus as much knowledge as possible about structure and pattern in the data before choosing methods to deal with the remaining uncertainty. This principle applies to new techniques like neural nets and fuzzy logic as well as more conventional mathematics.

Expert has 20 years of experience developing specialized optimization systems (incorporating specialized forecasting) for high-dollar-impact business processes where textbook answers are inadequate. His firm has developed software for large apparel manufacturers for material utilization, inventory management, orders from retailers, and store-level and aggregate retail sales. Their dynamic model stock management software recommends sizes, colors, or quantities of garments for each individual store. These systems give the effect of a "smart" system, but are based on mathematical analysis rather than AI techniques. The PC-based software is in ANSI C and uses Paradox(R)-format databases, but is designed for easy adaptation to different databases (UNIX, etc.)

Expert has 25 years experience developing specialized optimization systems (incorporating specialized forecasting) for large apparel manufacturers, including material utilization, inventory management, dynamic model stock management, analysis or forecasting of orders from retailers, and store-level and aggregate retail sales. Users get the effect of a "smart" system, but the logic is based on mathematical analysis rather than AI techniques. These systems deal with high-dollar-impact problems that don't have simple textbook solutions. Unlike linear programs, the bottom line for cut-order planning is not smooth. He has often found marker efficiency variation is not symmetric; it is easier to make a marker a lot worse than even a little better. Blips or "outliers" in sales for a stock keeping unit (SKU) can have a long aftereffect on moving averages or smoothing results.


Expert has 15 years of experience developing specialized optimization systems for large apparel manufacturers for inventory management based on analysis/forecasting of retailers' orders, and store-level and aggregate retail sales. "Need-to-make" analysis is based on comfort levels derived from fluctuation analysis. Intelligent Rounding(R) software plans production orders for garments so that the required quantity by size closely satisfies inventory requirements in a way that is compatible with making good cutting layouts and with efficient cutting and manufacturing. Dynamic model stock management software recommends sizes, colors, or quantities of garments for each individual store.


With his experience helping clients (typically apparel manufacturers) master interdisciplinary high-dollar-impact business processes, Expert has developed specialized optimization systems based on mathematical modelling of the process. Every time he has done a detailed study for a client, he has found surprises in the data that neither he nor the client expected, yet is significant for the project outcome. His firm therefore warns against defining and structuring a problem prematurely.


Expert has 25 years experience helping clients master interdisciplinary business processes that cross organizational lines. For example, analyzing, tracking, or forecasting orders and sales is an "intelligence" function that needs to integrate data from multiple sources throughout the company. Typically, responsibilities, ownership of data sources, and reporting relationships are fragmented, to the great frustration of user departments. Expert has found cases where business process re-engineering can be painless, not requiring radical changes in organization structure, but simply making operating procedures and systems smarter. His firm's cut order planning software relies on Intelligent Rounding(R) software upstream to satisfy inventory requirements while allowing good cutting layouts and efficient cutting and manufacturing.


APPAREL INDUSTRY. Expert led Systems Research at Levi Strauss & Co. for 12 years before establishing an independent business in 1985 to provide technical consulting and specialized number-crunching software. His clients have been major apparel firms including Levi Strauss & Co. and Haggar Apparel Co.



APPAREL PRODUCT RETAILING. Expert's firm has analyzed retailers' ordering and sales patterns for pants and other garments and has developed specialized analysis and forecasting techniques including season-to-date monitoring or assessment similar to election night analysis. The firm's dynamic model stock management software recommends sizes, colors, or quantities of garments for each individual store.