Building out developer relations and community development, particularly for the entity resolved knowledge graphs practice.

Managing Partner at the firm. Focused on open source integration for AI production use cases in enterprise: graphs, natural language, optimization, and so on. Applying theoretical and empirical tools to devise innovative solutions. A demonstrated ability to solve complex challenges. Addressing unique and difficult problems with major upside for clients. See https://derwen.ai/about

Technical and business advisor since the beginning; board member since 2022. Acquired by Hugging Face in June 2024.

Mostly as a subject matter expert, advising portfolio firms, assisting on some due diligence.

Our team helped introduce and further develop: Oriole (live coding tutorials based on Jupyter + Docker + video); the large-scale program for Live Online Training; formative self-assessments; and overall the structured "Learning Promise" metadata for learning materials. We developed ML apps for content parsing/classification, video transcription, recommendation systems, knowledge graph across content for 200+ publishers, etc.Assisted with conferences: Strata Data Conference, The AI Conf -- specifically for hosting Exec Briefings, conducting video interviews of sponsors, etc., and as program co-chair for JupyterCon. Co-chair for Ed Foo, 2017-2018.Gave public talks on behalf of O'Reilly Media, and taught a popular live online course, "Get Started with NLP in Python".

[+5 months earlier as consultant] Coordinating the Spark Camp events, creating a certification program in partnership with O'Reilly Media, developing instructional materials for Apache Spark, teaching, coordinating curricula, evangelizing through meetups/confs worldwide.

committer on the Cascading API open source project;domain expert helping guide Product and Engineering;developed sample apps and tests in Cascading, Scalding, Cascalog, PMML, R, etc.;developed machine learning libraries and PMML support;responsible for R&D of predictive modeling to analyze and manage large-scale workflows;responsible for internal analytics and reporting about customer engagement;technology evangelist and developer community manager;writing (O'Reilly, Hortonworks, etc.), webinars, and public speaking;responsible for business development and customer solutions;assisted with strategic partnering;recruiting/hiring manager.

Led team of Data Scientists and DevOps responsible for the data-related functions of the company, ranging from data warehouse (Hive on AWS) to metrics/reporting (Pentaho) to stats analysis (R) and machine learning (Hadoop) for marketing optimization, chat topic analysis, social recommenders, anti-fraud, etc.

Joined a small team to spearhead use of "Big Data" and machine learning within Jive SBS. Assisted on aggressive staffing effort to add ~100, mostly engineers.

Led efforts in machine learning and predictive analytics, for initial monetization of company data (3 yrs accumulated). Developed data infrastructure which was 100% in AWS cloud.

Built a large Data Science team, to recover failing recommender system which accounted for 80% of company revenue: 5 wks to hire team, 5 wks to recover production systems. Displaced $3MM capex in Netezza with $400/day Hadoop on AWS. Became largest Hadoop deployment on EC2, and case study for development of EMR service.

One of 4 founders in early-stage tech start-up backed by True Ventures. Text mining and NLP predictive analytics of social media. Initially focused on Second Life + Facebook integration. Led all technology efforts, moved company 100% into AWS when it launched. Managed teams in SLO and Russia.

Led a team of engineers in Austin and St Petersburg, Russia to develop UI for mainframe data archive product (metadata management) for Sarbanes-Oxley compliance.

Led engineering team to develop a Linux appliance for predictive analytics in network security: estimating risk metrics for network attacks in progress, and visualizing them in real-time. Runner-up for Apple Design Award, 2004. Installed at US Congress, Wildwell, etc.

Advised the exec mgmt and the S.P.E.C.T.R.E. cabinet on technical matters, in support of their quest for world domination. Assisted global operations teams – working under the code name "Dr. Xander" – in their assignments to eliminate Bond. Authored internal whitepaper to suggest novel methods for dispensing with evil henchmen who failed in their assigned tasks, and alternative methods for improving our shark feeding equipment. Introduced automated methodology practices for generating new industry buzzwords by applying ngram analysis to spam folders.

Austin, TX:Founded and ran early ecommerce firm (online bookstore + physical space) specializing in "long-tail", small press publications. Edited FringeWare Review magazine; received Wired Top 10 List 1993. Produced performance art events: Robert Anton Wilson, Negativland, Kenneth Anger, Survival Research Labs, Church of the SubGenius. Pioneered chatbots/avatars in customer support along with Robby Garner (open source as JFRED)

Acquired technology from Odin Corp. Commercialized machine learning apps for hardware-based pattern recognition, leading into industrial robots for automotive, etc.

Machine learning start-up (neural nets, fuzzy logic) for handwriting recognition in mobile apps, video recognition for military targeting systems. Technical liaison with Motorola. Supported legal team during lengthy intellectual property theft lawsuit in federal court (sole material witness in the trial), working with Scotty Baldwin.

Developed assemblers and cross-compilers for new microcontrollers, for Mac/Win/Unix. Supported sales engineering throughout Asia/Pac tour. Supported custom chip project for Toyota engine controllers. Assisted initial efforts to build chips for machine learning (neural net, fuzzy logic).

Hired into an expert systems group working on projects for telecom in Area 55. Later joined a neural network team led by John Hopfield in Area 11 (research), subsequently reorg'ed into Network Systems to help modify 5ESS (central office switch) to add X.25 integration. Later, developed one of the panels for the AT&T NOC to show a real-time map of nationwide alarms on AT&T fiber optic cables (Long Lines). Also founded the Mac User Group at Bell Labs.

Part-time research in Flight Guidance and Navigation branch for FAA project to build new generation of AI-assisted air traffic controller workstations.

Teaching Fellowship in CSD, to establish the CS1E course. Now known as Residential Computing. Recruited and managed team of 15 in peer instruction, wiring up campus dorms. Won a Dean's Service Award for this work.