PROFESSIONAL EXPERIENCE

Data Science, Machine Learnng, and Quantitative Analytics

Responsibilities:

Lately, I have been building an ICO server with React and MongoDB, but that is just a hobby. My professional work for several years has been building analytic tools for natural language and financial data, applying statistical and deep learning to predictive and causal modeling of socio - economic systems, to extract a latent model against which inference is applied. Before that I developed simulation modeling and adaptive integration systems. Previously I developed distributed systems, recommedation systems, network and systems software, and compilers

Adaptive Management Data Scientist

Responsibilities:

Devised predictive modeling strategies for corporate and sector business dynamics, applying standard R and Python packages, as well as purpose-built bayesian MCMC models written in Stan, to generate analytical predictive summary statistics and distribution envelope estimates for fundamental series and for factor models predicting revenue, and other business metrics. (Skills: R, Python, Flask, D3.js, Stan, C++, Celery, Jenkins, Jira, git, AWS, Redshift, PostgresQL, S3, data collection, ontology, management and analytics.)

Consulting

Responsibilities:

As a consultant for financial clients (market data, hedge fund, banking): Developed English-Japanese statistical translation system, textual open information extraction, question answering, dependency and semantic parsers, consumer credit risk and behavioral modeling platform and models, data warehouse integration schema, rate curve models, risk, liquidity and valuation metrics and predictors for interest rate products, trading strategy description language, simulation and back-testing systems, managed outsourced projects, designed and implemented a distributed simulation and analysis platform and framework, implemented textual provenance, market data and reference subsystems, storage hierarchy, dimensional slicing, caching, and streaming calculations, ported simulation packages and NLP pipeline to EC2. (Skills: C++/STL/boost, Python/Scipy/scikit-learn/statsmodels, Common Lisp, Javascript/jQuery, J6EE/JNI/JNDI/RMI/JDBC/JSP/Servlet/Swing, Perl/DBI/Twig, C#.Net 4.0, Windows, Linux, Solaris, AIX, PL/SQL, T-SQL, Fortran, MCMC (stan), (c/b)VAR, regression, neural networks, SVM, GPU/thrust/CUDA/Theano, SAS, UML, WSDL, FIX, TCP/UDP/IP/Multicast.) Project detail is found in the appendix, below.

Confidential

Senior Architect, Director of R&D

Responsibilities:

For defense applications: Invented natural and formal language semantic analysis and predictive state-machine generation technologies, engineered signal extraction from collected data, designed and developed adaptive middle-ware and virtualization run-time for simulation and enterprise integration with real-time message analysis and off-line ETL and ontological inference components, developed machine learning components and on-line learning architecture, presented results to clients, reverse-engineered legacy simulation applications, database and network data formats and protocols, designed and implemented automated ontology inference and alignment tools, managed federated simulation integration and interoperability events, developed server virtualization strategy, designed and implemented NIC firmware, RDMA drivers. (Skills: C++/STL Java/JNI, RDMA, Infiniband, Linux kernel drivers, RDFS/OWL, distributed processing, description logic, revision logic, genetic programming, LCS/XCS, Boosted Random Forests, Matlab, wxPython, Perl, Windows network drivers, hooking, VMWare, SQLite, MySQL, real-time message processing.)

Confidential

Software Engineer and Architect

Responsibilities:

For web media: Developed a web-based multimedia conferencing and collaboration system, with a scalable, redundant, low-latency message queuing system under real-time constraints, a redundant database back-end, and a mixed cross-platform client interface, developed enterprise security architecture and metrics, implemented development infrastructure and continuous-integration platform, managed engineering team, developed super-distribution platform for a content distribution network, designed and developed overlay network and content-based routing system, reverse-engineered and implemented network protocols for interoperability, designed clustering and associative data-mining components for recommendation systems. (Skills: C++/STL/ACE, reverse proxy, HA, MySQL clustering, Java2/JXTA/JINI/JGroups, DHTML/Ajax/Javascript, Perl, Linux, Windows, compression, MPEG, real-time,turbo codes, matrix completion, clustering, Bayesian regression)

Confidential

Computational Scientist, Staff Engineer

Responsibilities:

For scientific MPP application platforms: Developed scientific software (biochemical network models, meteorology and seismic codes, CFD, DSP, discrete event simulations, distributed join algorithms), developed operating system components, network drivers, hardware and software diagnostics, compiler modules and run-times, consulted for clients in academia, military intelligence, and petroleum, developed routing algorithms for distributed memory multiprocessors, optimized message-passing libraries and Infiniband drivers for throughput and latency, designed and implemented massively parallel I/O run-time and APIs, NAS management software, developed job and load management for HPC Clusterware, led and managed C* and HPF compiler projects over several release cycles, designed scientific libraries for distributed memory vector architectures, implemented Kerberos for HPC Clusterware, designed and implemented cluster management UI, performed usability studies on Java/Swing cluster management applications at the HCI lab. (Skills: C++/STL. Java/JDBC/JNDI/RMI/Swing/GSSAPI, MPI, MPI-IO, Infiniband, Verilog, Solaris kernel modules, Compilers, Lisp/CLOS, HPFortran, Fortran 90, Prolog, SAN, NAS, UX, ODE/PDE multigrid methods.)