WalmartLabs, Brisbane CA
Senior Software Engineer (Data Scientist), October 2013-present
– Data scientist in the multi-channel sourcing engine team. Developed the batch order optimization model and implemented it in Java, reduced total shipment zones by 14%.
– Developed MapReduce models for measuring outbound delivery performance, and used Cassandra API for data retrieving in the back-end system development.
– Lead data analyst for several key corporate initiatives, including Smart Ship-From-Store project (2014 Walmart.com CTO Award Winner), and Project Tahoe (walmrt.com prime membership program), etc.
– Tools: Java+Gurobi, Tableau, Hadoop/MapReduce, Hive, Cassandra
FedEx, Memphis, TN
Sr. Operations Research Analyst, July 2012-Oct 2013
– Leading air/freight/truck network optimization projects, in collaboration with various operating companies within FedEx.
– Participating in data analytics project to support corporate business initiatives.
– Received two Bravo Zulu Awards for significantly improving LPO optimization algorithm and driving efficiency in SQI process.
– Project highlights:
a. Leading developer and modeler of a linehaul plan optimization system. Applied OR knowledge to identify cost savings and profit improvement opportunities, designed algorithms for trip generation, dispatch generation, and optimization models, and implemented the system in Java+Gurobi, and MS SQL.
b. Leading developer and modeler for a facility location optimization system. Designed the facility selection algorithm, collaborated with business partners to model operating costs, and implemented the system in SAS.
c. Developer of an enterprise rail finder system. Developed the time feasibility module, rail schedule cost module, and opportunity matrix module.
d. Leading data analyst of a competitor business model estimation project. Designed advanced data models using SQL and optimization tools to predict competitor’s pick-up and delivery operations.
e. Leading data modeler of a China PUD improvement project. Designed regression models to analyze PUD cost factors.
JDA Software, Marietta, GA.
Operations Research Intern, June 2011 - Aug 2011
– Developed Oracle SQL packages to reconcile OD and Leg demand forecasters, manipulating and extracting information from very large, complex data sets.
– Modeled, prototyped, and developed Itinerary Demand Forecaster (large scale MILP problem), which has been integrated into the JDA cargo revenue optimization solution. It bridges OD and leg demand forecasters and feeds the itinerary demands into the arc allotment generator. This new feature has been used by Cathay Pacific Cargo in their 2012 cargo allotment planning (started in August 2011).
– Held regular meetings with the software development team in India and the client engagement team in Hong Kong.
– Experienced in:
Unix shell scripting + AMPL + Cplex + Oracle SQL Dev + SQL loader for prototyping, and;
Java + Cplex concert+ Oracle SQL Dev + JDBC for software development.
– Developed Oracle SQL packages to reconcile OD and Leg demand forecasters, manipulating and extracting information from very large, complex data sets.
– Modeled, prototyped, and developed Itinerary Demand Forecaster (large scale MILP problem), which has been integrated into the JDA cargo revenue optimization solution. It bridges OD and leg demand forecasters and feeds the itinerary demands into the arc allotment generator. This new feature has been used by Cathay Pacific Cargo in their 2012 cargo allotment planning (started in August 2011).
– Held regular meetings with the software development team in India and the client engagement team in Hong Kong.
– Experienced in:
Unix shell scripting + AMPL + Cplex + Oracle SQL Dev + SQL loader for prototyping, and;
Java + Cplex concert+ Oracle SQL Dev + JDBC for software development.
Webco Industries, Sand Springs, OK
Co-op Operations Research System Developer, Jan 2010-June 2011
– Collaborated with Webco’s management team to translate their business requirements into structured optimization models that enhanced firm wide decision making; The Decision Support System is connected to Webco’s SAP system by the database interface.
– Comparing with historical data, resource utilization was improved by 14%. This improvement translates into a reduction in waste, holding cost and variable production costs (e.g., wrapping and bundling cost). Overall due date violation was reduced by about 18%. The total inventory holding cost was reduced by 2.3%.
– Developed 14 Available-to-Promise Mixed Integer Linear Programming (MILP) models to reduce inventory-sales mismatch by dynamic allocation of inventory/purchase orders and efficient scheduling of production capacities.
– Significantly reduced the available-to-promise quoting time for the sales team while accommodating new customer orders.
– Provided options for bumping/delaying existing sales orders to satisfy high premium orders/high priority customers.
– Experienced in:
Xpress MP for prototyping, and;
C++ with Cplex concert and Gurobi for optimization engine development.
– Collaborated with Webco’s management team to translate their business requirements into structured optimization models that enhanced firm wide decision making; The Decision Support System is connected to Webco’s SAP system by the database interface.
– Comparing with historical data, resource utilization was improved by 14%. This improvement translates into a reduction in waste, holding cost and variable production costs (e.g., wrapping and bundling cost). Overall due date violation was reduced by about 18%. The total inventory holding cost was reduced by 2.3%.
– Developed 14 Available-to-Promise Mixed Integer Linear Programming (MILP) models to reduce inventory-sales mismatch by dynamic allocation of inventory/purchase orders and efficient scheduling of production capacities.
– Significantly reduced the available-to-promise quoting time for the sales team while accommodating new customer orders.
– Provided options for bumping/delaying existing sales orders to satisfy high premium orders/high priority customers.
– Experienced in:
Xpress MP for prototyping, and;
C++ with Cplex concert and Gurobi for optimization engine development.
Center for Computer Integrated Manufacturing Enterprises, Oklahoma State University, Stillwater, OK
Research Assistant, Jan 2007-July 2012
“Sales Effort Free Riding and Coordination with Price Match and Selective Target Rebate," supported by Oklahoma State University.
– This project studies sales effort coordination for a supply chain with one manufacturer and two retail channels, where an online retailer offers a lower price and free-rides a traditional retailer’s sales efforts. The free riding effect reduces the traditional retailer’s desired effort level, and thus hurts the manufacturer’s profits and the overall supply chain performance. To achieve supply chain coordination, we designed a contract with price match and selective target rebate. We showed that such a contract can arbitrarily split the supply chain system profit and achieve Pareto improvement.
– Experienced in:
Game theory and stochastic nonlinear programming for modeling and
Matlab for numerical analysis.
“Sales Effort Free Riding and Coordination with Price Match and Selective Target Rebate," supported by Oklahoma State University.
– This project studies sales effort coordination for a supply chain with one manufacturer and two retail channels, where an online retailer offers a lower price and free-rides a traditional retailer’s sales efforts. The free riding effect reduces the traditional retailer’s desired effort level, and thus hurts the manufacturer’s profits and the overall supply chain performance. To achieve supply chain coordination, we designed a contract with price match and selective target rebate. We showed that such a contract can arbitrarily split the supply chain system profit and achieve Pareto improvement.
– Experienced in:
Game theory and stochastic nonlinear programming for modeling and
Matlab for numerical analysis.
_EMC&RF Lab of BUAA, Beijing, China.
_Research Assistant-Pattern Recognition, June 2005-June 2006
– Developed feature identification and pattern recognition algorithms for radar signal processing by using Matlab and C++.
– Developed an EMC expert system for airplane design by using VC++, VB, SQL2000 and Matlab, featuring 3D modeling, behavioral simulation and hardware-in-the-loop simulation technology.
– Developed feature identification and pattern recognition algorithms for radar signal processing by using Matlab and C++.
– Developed an EMC expert system for airplane design by using VC++, VB, SQL2000 and Matlab, featuring 3D modeling, behavioral simulation and hardware-in-the-loop simulation technology.