Refactor BioSciences worked with the Sexton Lab at U. Michigan to identify early stage inhibitors of COVID-19

I currently lead the software development and scientific informatics teams at Translucence BioSystems. In addition to helping managing the informatics operations for next generation histology pipelines, the group also leverages AI/ML solutions for the quantification of multi-terabyte sized 3D Lightsheet datasets, enabling companies to quantify differences in proteomic changes in specific regions of the brain, as a result of a peturbation such as a drug or knockout.

Juva Health is a digital health company that aims to be the best non-drug treatment addressing the world’s most prevalent health problems: Migraine, Chronic Pain and Anxiety. We uniquely offer smartphone-based biofeedback using a remote-sensing technology as an alternative to clinic-based treatment. Now patients can easily access a treatment modality that’s scientifically proven to be as effective as medicine and is offered by many hospitals and reimbursed by many insurance carriers.

Shop My Porch is a platform that connects buyers who are fanatic about supporting local makers. We highly encourage truly shopping local - we help buyers discover makers in their own backyard, encourage building communities through supporting one another, and enable makers to flourish through fee-free selling.We have built a flutter based app integrated with firebase / django heroku website backend, supporting over 20K users, hyperlocalized in the San Diego Area. Visit https://www.shopmyporch.com to learn more.

Collaborative Drug Design is an informatics company specializing in tools for drug discovery. I am currently working on generative models for chemistry using deep learning technologies. By combining graph convolutional neural networks and GRUs, we have developed a patent-pending generative model that can be used to recommend compounds with improved chemical properties. (See https://optibrium.com/events/webinars/webinar-perspectives-on-generative-chemistry-potential-and-reality/ for more detailed information). Prior to this project, I also worked on the BioAssay Express product – a tool that leverages NLP for converting free form text into searchable metadata. (https://www.bioassayexpress.com).During this period, I have helped improve user interfaces; implement a selenium testing framework; implement LDAP authentication; and led the integration efforts for one of the BioAssay Express customers.

As a part of team of drug discovery specialists, I work together with other experts to move drug discovery pipelines forward.My current areas of specialization include:- Data Mining/Integration- Bioinformatics & ChemInformatics- Target identification- Drug repositioning- High throughput screening systems- Systems/pathway biology- Machine learning- Text mining- Natural language processing

Co-Founded with my wife Phyllis Cheung, LuxeFinds and its disruptive color matching technology has been featured in The Wall Street Journal, Entrepreneur, Inc., Glamour, CBS News, ABC, Cool Hunting, and Apartment Therapy. Leveraging my experience in software architecture and my data sciences background, we built LuxeFinds.com, a unique shop by color experience which allows users to search by exact shade of color across 1.5 million products in near real time. We auto-tag categories using ML text classifiers and leverage graph data / network experience to identify core hub products for consumer stratified groups. We use this data internally to determine what new products to promote on our Pinterest account which gets 9 Million exposures a month.

Served as the interim CTO for Brighter.do -- SAAS company focused on helping customers create a company culture that engages employees and makes top performers want to stay. I was brought on board to increase scalability in the platform. Re-architectured core design to enable messaging system to scale to new customers on IOS and Android clients. Involved changes in core service layers and database design.

A small group of employees were retained to help support assets marketing efforts. During that period, a two teams were formed – a primary team focused on large market indications, and a smaller team focused on repositioning lower value assets for orphan / rare diseases. I led the secondary team and developed the technology and business plans required to reposition DNS assets into orphan and rare diseases. To reposition drugs, I designed a disease graph engine that leveraged machine learning to build relationships between rare diseases and DNS assets.

During my time at Dart NeuroScience, I built and grew the Bioinformatics / Computational Biology Group at Dart NeuroScience to provide solutions to the challenges presented by the drug discovery process. Supported the early target discovery Pipeline Thru 2021: Built/design workflows, informatics tools, and requisite infrastructure to empower teams to fill early target discovery pipeline thru 2021. Leveraged integrative analytics and data science for novel target discovery: Built an integrative omics graph which integrated all of the data from DNS experiments with public datasets. Built network visualization tools and implemented network analysis methods. Leveraged machine learning approaches leveraged machine learning to uncover new hypothesis and increase signal. Used text mining and natural language processing (NLP) to mine the literature to identify papers linking targets to disease.Researched and designed novel algorithms: In addition to implementing standard NGS techniques, also implemented custom in-house data analysis pipelines. I was granted US Patent US8868349B2 for microarray analysis techniques, which has been since adapted to NGS differential expression. Several patents filed for in-house behavioral testing software / hardware. Collaborated with biologists on advanced experimental techniques.Active project team participation: Active contributing member of the Candidate Memory Gene Identification and Target Evaluation teams. The Candidate Memory Gene Identification team reviews all data generates lists of targets to be reviewed/prioritized by the Target Evaluation team Strategic Projects / Cross-Departmental Innovation: The bioinformatics group acts as an innovation group for addressing problems outside of the traditional bioinformatics realm. Impacted Cognitive Sciences, Pre-Clinical, and Clinical groups with projects including a software / hardware solution that used at DNS’s Florida Cognitive Research Unit in human clinical trials.

Indication Discovery Platform: The Target Prospector Team Network (TPTN) an innovative software platform for fast-tracking new target discovery in Opthalmology Therapeutic Area called The framework combines a variety of data sources from gene expression to Natural Language Processing (NLP) text analytics to help prioritze targets for review. The key differentating factor from the current state of the art includes an user annotation / collaboration engine which allows the team to quickly work together to identify the top targets. Computational Biology Project Support: Provided support on early stage CXCR4 program. Performed CXCR4 pathway analysis and diseases stratification studies based upon gene expression found in cell lines. Program eventually resulted in PF-06747143, a biological currently in Phase I clinical studies.

Lead a team of developers and curators to successfully deploy a suite of applications for accelerating Structure Based Drug Discovery across Pfizer’s seven discovery hubs.Structure Based Drug Design Informatics: Software Technical Lead for team development of Pfizer’s Advanced Drug Discovery System (ADDS). Globally deployed throughout Pfizer, ADDS is the repository for all crystal structures solved at Pfizer and relevant public structures. The system automatically processes crystallographic data and performs asymetric unit (AU) extraction, appropriate side chain and hydrogen modeling, and superposition into a common reference frame. Additional data such as electron density maps are also calculated. These are accessible from ADDS Explorer, a SAR-style tool which allows scientists to compare assay results against compound / crystallographic properties (since replaced by D360). Finally, there is a service layer integration with MoViT, Pfizer’s SBDD visualization / design tool, which allows scientists to quickly access crystal structures that are ready for computational design.

Technical Lead for development of supply chain solutions for the medical industry. Worked on ERP (JD Edwards) integration with our suite of web based products (auction, reverse auction, rfp, etc...) and Commerce One products.Primary duty: Technical lead and design architect for new Java J2EE registration/profile process. New registration process involved design of new business objects, component based JSP MVC model. Wrote functional specifications and architecture documents. Managed team of developers to deliver product on time. Utilized WebLogic 5.1 Application Server. Database interface layers task included transitioning homegrown portal to Commerce One MarketSite technology. Moving 10,000 users from Oracle 8i to LDAP.Secondary duties: Developed Second Generation E-Commerce site – primarily focused on eSpecials supplier side module. Wrote JDBC ERP Integration class with JDEdwards OneWorld application. Utilized Netscape web server and Jrun servlet engines. Oracle 8.1.5 database with SQLJ database layer. Proponent of Weblogic 5.1 migration and core service standardization.

Designed and implemented event management system that was deployed in the "War Room" at WorldCom/MCI, Mcleod USA, and Teleway Japan.Primary Duty: Technical lead and core developer for a telecommunications software company. Project involved the design and implementation of the InvenSys Enterprise Database System (IEDS) and the Event Management System (EMS). Created requirement, design and testing documents for an offshore development team located in the Cayman Islands, while working closely with onshore staff to design IEDS core database schema. The project utilized a client/server architecture using RMI. The multithreaded server utilized a JDBC connection to an Oracle 8 database. The client was designed with Swing 1.1, supported internationalization, and was deployed in Japan with kanji user interface. Secondary Duties: Alarm Resolution Database (ARD). The ARD is a helpdesk/training application written in Java; ARD communicated with both Java and OpenStep applications over sockets, and utilizes JDBC to communicate with any ODBC compliant database.Other previous projects: Porting the Circuit Alarm Server and Circuit Alarm View applications from NextStep 3.3 to OpenStep 4.2, as well as redesigning and implementing new features. Other projects include developing a GUI application to parse and edit configuration files, porting and maintaining GUI application insepectors, the development of a Web based bug tracking syste, and a metrics analysis of distributed objects technologies (DO, RMI)

Technology Research Group – Laboratory Sensors and Automations Division. Core developer for a robotics software package. Allekto System primary product involved a runtime dynamic high throughput screening robotics system for application in drug development and combinatorial chemistry. Primary responsibility involved work unit software design and implementation, system setup/integration (Hardware and Software) and system testing. Secondary responsibilities included device driver work as well network support.