I do advanced data science R&D and software design, for a property insurance MGA startup firm. I have developed new pricing technologies that boost profitability and optimize risk exposure. I measure and analyze the impact of insurance risk selection. I advise on architecture for software testing and reliability. I have designed and built an innovative explainable AI (XAI) system for insurance catastrophe risk modeling, on hurricane, tornado, wind storm, hail, and other perils – using machine learning (ML) and artificial intelligence (AI) analytics.

I founded WRS.ai, a company for my independent consulting work. I specialize in design, development, and deployment of smart software technologies. I finished major contracts ahead of schedule and within budget – for short- and long-term clients and projects.(Independent consultant.)

I do business and technology guidance for early‑ and seed‑stage startup companies. Clients have included: • LULU'S (artisanal cafe), NJ. Strategy analysis for retail branding, marketing, PR. • SIRCH (Internet search engine), NJ. Business models, products, AI technology. • LAVAMAP (real estate search engine), NY. Business models, software architecture, AI. • LULUCHAIN (blockchain), NY. Distributed ledger technology, cryptocurrency, ecommerce. • DEVOTED.AI (ecommerce), CA. Artificial intelligence, system architecture, sales optimization. • LUNETO (medical errors prevention), NY. Electronic health records, legaltech, NLP.

I upgraded the software cybersecurity of NYU's central global identity-management system for students, faculty, employees, affiliates, and alumni. I created custom software tools to detect and repair security vulnerabilities. I hardened NYU's worldwide system for logins and access, by rewriting code, tightening webserver configurations, reducing attack surface exposure, upgrading encryption, and adding web application firewall.(Major university identity and access management – IAM.)

I built software for news aggregation, including:• AI to identify in real time the hottest news: TOPNEWS.• Machine learning to predict news impact on stock price: IMPACT.• Algorithm to find true jumps in stock price time-series: SCRUNCH.• News topic clustering, deduping. US Patent 9,753,964: MIMOSA.• Optimally combine classifiers to reduce errors: ENSEMBLE.• Scraper to isolate news content: DECHROMER.• Combinatorial-optimized bidding for content: AUCTION.• Text-pattern compiler for low-latency stock trades: FLASHBANG.• Spider for SEC business filing documents: EDGAR.• Data analytics engine for sales, marketing, editorial: TALLYHO.• Lemmatization dictionary for search and clustering: LEMME.• Content signatures to classify and summarize articles: FACET.• Improved topic taxonomy, entity disambiguation: METABOT.• Extract meaningful tweets from full firehose: TWITTER.• Automatic data GUI, workflow, drilldown: SCREENTEST.• History-tuned monitor for newsfeed outages: WIRETAP2.I held multiple leadership roles, with responsibility for product/business strategy and planning, product architecture and design of natural-language processing software, big data analytics, data science, semantics applied research, customer liaison, patent writing, technical staff recruiting, rapid hands‑on implementation, and project management. I led my team to upgrade software engineering processes for agile development, testing, and reliable deployment. AM builds and runs advanced software products that analyze, sift, cull, classify, aggregate, and deliver news. Over 500,000 news articles per day, from thousands of publishers, flow through AM's high‑speed natural-language processing engine, which automatically identifies topics, people, companies, locations, events, relevance, and predicted impact. AM indexes and delivers articles to syndication customers based on their interests. Customers include financial brokerages, stock traders, and corporate researchers.

I specialized in performance enhancement and optimization of our enterprise-grade website back-end software. I co‑led our system architecture initiative for multi-datacenter scaling. By introducing a template compilation system, I achieved a 4x load reduction across the Hearst webserver farm. In addition, I reduced database replication lag across datacenters. I increased the reliability and flexibility of our software testing and content management infrastructure. I developed and deployed new social/community features for the magazines.Our 15‑person team designed, built, and operated the software to create, publish, and run 48 major magazine websites, including Oprah, Cosmopolitan, Good Housekeeping, and Esquire, with millions of viewers. Our software content-management system (CMS) platform supported authors, editors, graphic designers, photographers, and advertisers, to create and deliver every article across all publications.

I was in charge of the entire startup product, technology, and team. I helped originate and co‑develop the product vision, design, and business strategy. I won investor funding commitment. I recruited, hired, mentored, and taught technical and creative staff – tripling team size, to 12.I led all team operations and decisions: hiring, resource planning, marketing, agile full-stack software development, content, user experience design, visual design, platform architecture, scalability, database, languages and frameworks, APIs, analytics, tech support, and open‑source licensing.I managed the LimeBits.com alpha consumer launch, with our suite of 45 mixable website designs and apps with 1‑click copying, for blogging, social networking, wiki, invitations, scrapbooking. Gave public talks and wrote whitepaper. Grew a community of users.LimeBits was a startup venture -- we built an innovative open-source community platform for sharing website designs and features. We made it easy to create your own website by copying and modifying the shared structure from existing sites, rather than building from scratch.

I built parts of the consumer-facing website, including HIPAA security, reliability, reports, and admin system.(Quintiles drug safety alerts startup.)

I helped assess startup venture capital deal proposals. I led partners through the deployment of a customer relationship management (CRM) and knowledge management (KM) system. I liaised with and supervised an overseas software development team. I managed the firm's IT systems, including network, VPN, email, website, CRM, and KM.(Wireless telecom venture capital firm – VC.)

I built and tuned an algorithmic stock market crossing network for optimally executing equity trades to gain liquidity while minimizing market impact. I mathematically analyzed price risk in trade executions. Team project planning; patent writing.(Stock brokerage startup.)

I researched and advised on software startup feasibility.(Tech venture capital investor – VC.)

I designed and built a fast website search engine and navigation system.(Lifestyles popular publication.)

I designed and built tools for wrangling, extracting, cleaning, analyzing, modeling, and reporting financial data for equity/stock traders and researchers. I created a system for deploying and running distributed software in the cloud or grid.(Investment bank.)

I designed and built novel software for high-precision search, web-scale spidering, text indexing, and consumer-facing graphical user-interface (GUI).I co-founded the company with two colleagues, and led investment fundraising efforts.(Web search engine startup.)

I built a configuration-management system for an ecommerce website, to manage security, scaling, and replication.(Children's educational book publisher.)

I built and deployed a scalable web privacy service. I collaborated with team members on product strategy, marketing, and corporate acquisitions.(Nationwide ISP startup.)

I built a full-stack platform for interest-based social network communities, using member-created questionnaires and polls.(Social networking startup.)

I did rapid full-stack software development of a web privacy service derived from Bell Labs technology. I implemented encryption, antispam, security, identity, and automation features. I did market analysis, and I prepared and performed investor pitches with our CEO.(Lucent Technologies web privacy startup.)

I created and evaluated novel computer vision algorithms for human face identity recognition, compression, and search. I developed visual face recognition software, for award-winning biometric products. I served as project manager and primary liaison for a key customer. I coordinated activities of our 8-member R&D team.(Biometric visual face recognition startup.)

I performed and led multidisciplinary academic research at the intersection of neurophysiology, psychology, vision, and computation. I invented algorithms in neural computation, neural networks, machine learning (ML), artificial intelligence (AI), computer vision, math, statistics, data analysis, clustering, and pattern recognition. I built simulation models of complex systems. I built software for computer graphics, data visualization, and graphical user interfaces (GUI). I designed and ran human visual perception experiments.• I founded BrainLab, a research lab in which I hired, trained, funded, and directed postdocs and grad students. I advised 16 PhD dissertations and 9 MS theses.• I served as Principal Investigator on grant-funded research, and received awards of $1.1M, from agencies including the National Institutes of Health (NIH Shannon Award), and the Office of Naval Research (ONR). I served as co‑investigator on grant-funded research projects receiving $69M in funding.• I taught graduate courses in neural computation, visual perception, and cognitive science, and undergrad courses in programming languages, object-oriented software, and digital electronic logic.• I served on academic journals, conferences, and funding agencies.• I authored/coauthored 50+ scientific articles, published in NIPS, Neural Networks, Network:CNS, SPIE, IJNS, IJCNN, INNC, Neurocomputing, Visual Neuroscience, ARVO/IOVS, JOSA‑A.• I held adjunct faculty roles at UNC and Duke Univ in Cognitive Psychology, Cognitive Science, Biomedical Engineering, Neuroscience, and Neurobiology, in addition to Computer Science.

I performed scientific research on visual perception and neural computation. I developed and ran simulation models of adaptive neural network processes in vision, using vector supercomputers to numerically solve complex, massive systems of coupled nonlinear differential equations.(Center for Research in Learning, Perception, and Cognition / Minnesota Supercomputer Institute.)

1988 to PRESENTSoftware architecture, news content analysis, full‑stack software development, algorithm design, cloud architecture, content management (CMS), customer relationship management (CRM), relational databases, blockchain, Software as a Service (SaaS), Software Development Life Cycle (SDLC).Software engineering, research & development, agile, scrum, DevOps, CI/CD, Docker, Kubernetes, A/B testing, ITIL, B2B, B2C.Data science, machine learning (ML), artificial intelligence (AI), neural networks (NN), RNN, CNN, deep learning, cognitive computing, decision trees, data similarity clustering, regression, PCA, SVM, regularization, bagging, boosting, reinforcement learning, unsupervised/supervised learning, ensembles, risk modeling, time series, attribution, predictive analytics, statistics.Natural-language processing (NLP), entity extraction, disambiguation, semantic tagging, topic modeling, text/news classification.Data visualization, data cleaning, ETL, data governance, graphical user interfaces (GUI), user experience (UX) design, information architecture, graph algorithms, Neo4j, trading signals, FLOSS, open source, libre software.Digital strategy, cross-functional team leadership, web services.Modeling, simulation, pattern recognition, computer vision, image processing, signal processing, high frequency algorithmic trading, optimization, object-oriented software, knowledge management, search engine internals, privacy/security.Python, R, JavaScript, Node.js, Perl, PHP, C, C++, Java, Lisp, lex, shell, XML.Dataiku, CART, C5, NumPy, SciPy, Pandas, Word2vec, ARMA, ARIMA, GARCH, Matlab, RDF, Spark, Hadoop, Pig, NLTK, Stanford NER.Linux, Unix, Apache, Tomcat, Elastic, Lucene, LAMP, MEAN, DOM, SQL, MySQL, Postgres, PostGIS, Sybase, Oracle, MS SQL Server, noSQL, Redis, Windows Server.Google Cloud (GCP), Amazon Web Services (AWS), Git, Subversion, WordPress, FIX, WebDAV, REST, RESTful API, MVC, Nagios.