ChemImage is a scientific company that applies a core technology of chemical imaging via hardware and software to solve problems for customers in a diverse range of applications such as threat detection, forensics, pharmaceuticals, semiconductors, and the biomedical field. Chemical imaging can be applied to any application in which there is a heterogeneous distribution of materials at a given spatial resolution – both microscopic and macroscopic. The algorithms developed take advantage of spectral, spatial, and temporal information.• Contributed to the development of algorithms in Matlab and C#. The emphasis has been on image processing, computer vision, and machine learning algorithms for specific projects and products• Responsible for the management of a group of six algorithm scientists• Supported and attended multiple field tests for the Threat Detection line of business• Principal investigator for the Real Time Chemometrics and Sensor Fusion (RTCSF) SBIRs (2/6/8 through 2/6/11) o The Real Time Toolkit (RTTK) was a key outcome of this project and was the result of a collaborative effort with the software group. o A further outcome from this project was the “award of a $17-million contract by the U.S. Army Space & Missile Defense Command for the implementation of its real-time sensor technology that will be used in the field to detect explosive threats” (press release 2/28/11).• Principal investigator for a Commercialization Pilot Program (CPP) that is a follow-on to the RTCSF SBIRs with the goal of porting the automated image analysis algorithms to an embedded processor (9/1/11 through present)• Contributed to the company’s aggressive patent portfolio development through submitted, allowed, and assigned patents• Have presented multiple papers at conferences to support the company’s presence in the scientific community

The core ChemImage software product, ChemImage Xpert, is a platform product that consists of approximately 500,000 lines of code and provides for the acquisition of chemical images from ChemImage hardware and the analysis of those images using digital image analysis and multivariate statistical techniques.• Participated in the day-to-day management of a group of five software developers including project planning and scheduling, technical guidance, adherence to good software development lifecycle practices, hiring, customer support, testing, and documentation • Led the technical efforts to maintain and improve the ChemImage software products. • Worked with the group on a complete redesign and rewrite of the software from a Borland C++ development environment to the Microsoft .Net development environment using C#• Worked on numerous modules of the software throughout all phases of the SDLC including SRS documents, architectural documents, and coding and testing• Follow the philosophy that bugs should be avoided by painstaking attention to detail in all aspects of the SDLC and by completely understanding the code that is being written• Have been involved in a variety of algorithm development projects including an algorithm for the automated detection of defects in semiconductor wafers, an automated baseline correction algorithm, and an automated spectral calibration algorithm• Was the Principal Investigator for an FBI-sponsored project – “Forensics Integrated Search Technology (FIST)” from 10/19/05 – 1/31/07• Contributed to the company’s aggressive patent portfolio development through submitted, allowed, and assigned patents• Have presented multiple papers at conferences to support the company’s presence in the scientific community

• Development of software in C++ for the simulation of mass spectra through the use of fragmentation rules• Creation of chemical structure handling routines including determination of equivalency of chemical structures and structure traversals• Identification of incorrect spectra

• Application of QSPR techniques for the prediction of dielectric constants • Development of a database of 800 compounds with known dielectric constants• Development of programs in C for the calculation of molecular descriptors• Development of neural network software using the BFGS training algorithm• Use of neural network techniques for model building• Use of Gaussian 94 for the calculation of dipole moments, polarizabilities, and partial atomic charges• Use of data analysis techniques including PCA and cluster analysis• Implementation of the developed computer code on a cluster of SGI Origin 2000s

• Thesis entitled “Automated Carbon-13 Nuclear Magnetic Resonance Spectrum Simulation Based on Database Direct Retrieval and Model Building Techniques"• Automated molecular modeling for all structures in the Sadtler database using MM2• Use of parallel programming techniques to parallelize the MM2 code• Creation of graphical chemical structure display routines• Creation of chemical structure handling routines including ring finding and structure traversals• Development of multi-dimensional partitioning code making extensive use of recursion and pointer-based data structures such as linked lists• Identification of incorrect spectra as outliers• Enhancement of chemical environment encoding algorithm• Use of PLSR, PCR, and MLR for model building• Extensive programming in FORTRAN and C on UNIX systems

• Analysis of paints and coatings for composition• Use of GC, IR, AA, HPLC, GPC, UV, titrimetry and other wet methods• Computer applications including a user interface for project reports using dBase III and C; an X-ray diffraction interpretation program using LISP; various paint calculation programs