TECHNICAL SKILLS

• SAS Base
• Stats ETS in Windows
• UNIX and mainframe. SQL Database
• Oracle and DB2. Create and update databases wif millions of records hundreds of tables and frequent updates. All MS Office products.

PROFESSIONAL EXPERIENCE

Software Engineer, Product Manager

Confidential, Glen Mills, PA

Responsibilities:

• Design, program, pricing and strategy of commercial trading subscription application for electricity markets. Application covered PJM, the largest power market in North America and price forecast hundreds of locations.
• Large multi - variable SAS regression models from millions of records wif SAS ACCESS to SQL Databases. Extensive use of SAS Statistics, SAS ETS, and macro language to organize and process thousands of models. Creation of point/click reports to surface wide array of trade options relative to client risk/reward preferences. Compose extremely accurate method for price estimations in real-time when power pool does not publish.
• Data collection SAS programs to obtain external website information and maintain large SAS databases.
• Creation of hydro dispatch model to optimize power generation wif water inflow, outflow, and storage. Composed multiple models for AES Energy and Covanta to optimize diverse electricity power plant types into day ahead and real time markets. Personally conducted many hour-long application demonstrations and training.

Confidential, PA

Senior Energy Trader

Responsibilities:

• Electricity commodity trading in Eastern U.S. power markets using ICE financial swaps and PJM Inc/Dec market. Inc/Dec market is where thousands of bids and offers are placed in blind auction day in advance for next day. Numerous and complex price forecast models and auto-generated bid and offer routines using SAS.
• Entire automation of load serving program for numerous metro areas and towns to serve customer load groups. Optimized statistical routines estimated the max day ahead price to buy, otherwise purchases made in real time. Construct large semi-normalized SAS databases wif multi-year millions of records and one-minute updates. Strategy programs and automated power pool electronic submittal routines programmed in SAS and Java... Forecasting of electricity load for Penn State University, Amtrak, New Jersey municipalities of South River, Park Ridge, Seaside Heights, and Lavallette. Use of weather data and multiple SAS products to program.
• Models remained in place wif former employer for over 5 years.
• Complete automation of PJM wholesale energy usage weekly billing components for clients.
• Create complex market reports of generation fleet status, transmission system, and spot natural gas prices.
• Serve as key quant between Risk Group, Trade Floor, and Management.
• $2 million average annual net trading returns achieved while Confidential over 5 year window.
• $1 million made on monthly load hedge between Dominion and Allegheny Energy.
• Daily Inc/Dec trading, Monthly FTR trading, Annual FTR trading and all short-term ICE Swap trading products. Due diligence financial valuations generated for power plant asset purchase and sales.
• Special complex analysis performed on Bath County pumped storage hydro plant dat involved 3,000 MWs of power dat could be consumed at well as generated wifin a 20-minute notification window.
• Perform valuation routines on customer load serving aggregates for utility zones in the U.S. Northeast.
• Due diligence valuations conducted for several New England pumped storage power generation plants. Virginia generation fleet dispatch modeling in SAS wif 20 weather year simulations for stress tests.
• Key analysis for Confidential ’s bid for all of Niagara Mohawk’s New York generation fleet.
• Strategy formulation regarding several combustion turbines constructed in Western PA.
• Analyze the nodal PJM Inc/Dec market initiated in June 2000 providing trade recommendations to trade floor.. Became a PJM trader by 2002 coz of the opportunities uncovered in dis new market.
• Identified trends, patterns, and generated complex SAS programs to implement profitable trade strategies.
• The PJM market opened initially wif 5,000 nodal locations and the 24 individual daily hours as trade options.

Database and Marketing Analyst

Confidential

Responsibilities:

• Multi-million dollar savings proposal for scaling back oversized Virginia residential transformers.
• Market assessment of Northeast retail electricity load serving based on newly developed wholesale market.
• Full programming of and marketing analysis for pilot program of surge protection devices in Virginia statewide. dis involved a 200-variable exchange between telemarketing, billing center, and installation companies. . Successful pilot turned into full business rollout, generating an entire new product offering by Virginia Power. . Complete construction of a 1.8 million residential customer database in SAS for Dominion Virginia Power. Performed rigorous customer segmentation analysis and management decision support.
• Programmatic preparation of monthly volume of 1 million direct mail solicitations for banking product offerings. Complete automation of electronic lists selection and output to numerous mailing vendors wif various formats. Developed name and address key process to handle all monthly mailings prior to purchase of Postalsoft.
• SAS used to interface wif ORACLE database of 50,000,000 consumers via SAS Access to Oracle.
• Performed response rate analysis on two-year marketing program wif zip-code level analysis.
• dis analysis generated best response rates ever achieved in the marketing group for the most profitable product. . Analysis of Signet Bank 800 phone center and identification of customers who cost the bank $1 million annually. Formal analysis of 1 million credit card accounts partitioned between Confidential and Signet Bank.
• Identified multi-level corporate bias in Signet Bank receiving low-performing accounts from Confidential .

Senior Risk Analyst

Confidential, Richmond, VA

Responsibilities:

• Composition of full simulation model to quantify legislative impact costs for Virginia’s inmate population.
• Monte Carlo simulation model data preparation and running of 10-year prison system forecast.
• Create forecast of state parole population formally selected through multi-agency submit and review process. Identified major error in the “worst offense” algorithm previously used for inmate institutional assignments.
• dis lead to a major change in the official programmatic agency classification of inmates.
• Automation of complex census tract level unemployment rate project to maximize federal dollars to Virginia. Compose entirely new accepted model to replace federal-mandated unemployment rates sampling process in VA. Rigorous occupational surveys of 10 page and 30 page surveys for statewide estimates of labor force makeup. Credited wif obtaining VA’s largest employer of 38,000 employees to respond for the first time to dis program. Debug federal occupations regression projection program used throughout the country.
• dis was done by identification of a FORTRAN and math error in the Gram-Schmidt Orthonormalization routine.