Founder at N54, Inc.
Greater Los Angeles Area
Founder at N54, Inc.
Greater Los Angeles Area
I enjoy leading moderate sized teams focused on prototyping, speculative development, and translation of research into development ("high risk, high reward" activity). I like iterative test-driven development, data-driven decision making, and positive psychology oriented management.
Soft real-time high volume server engineering; real-time and near-time decision system engineering; machine learning, pattern matching, and data mining; web monetization.
(Internet industry)
January 2008 — June 2009 (1 year 6 months)
N54 was small company (5 engineers) focused on providing artificial intelligence solutions for content monetization to ad networks and large publishers. Our technology was implemented in Erlang (server-side) and Javascript (client-side), and made extensive use of Amazon’s EC2 and S3 services. We also utilized scalable on-line learning algorithms resulting in optimized decisions delivered in under 400ms at scale. Unfortunately we were unable to generate sufficient traction with our customer base to motivate follow-on investment.
(Public Company; 1001-5000 employees; YHOO; Internet industry)
January 2004 — November 2007 (3 years 11 months)
Highlights:
* Built initial YPN prototype: innovations include response modeling impacting ad selection; eliminating Oracle from serving system; message queue based data replication; integrated revision control, packaging, and build environment system for increased developer productivity.
* Led team of 9 engineers focused on optimization, response prediction, and marketplace design.
(Public Company; 201-500 employees; Internet industry)
April 2003 — December 2003 (9 months)
* Designed and built initial (non-editorial) Domain Match prototype after ASI deal collapsed; product is now a significant revenue source for Yahoo.
* Built initial system for optimizing north inventory utilization for sponsored search.
(Privately Held; 1-10 employees; Biotechnology industry)
January 2001 — April 2003 (2 years 4 months)
Started a biotech company with my friend Albert Erives around the concept of regulatory DNA identification (by software) and manipulation (by microbiology) for purposes of drug discovery. Unfortunately it was a challenging period to raise money, and we eventually went broke. On the plus side, I learned alot about practical aspects of entrepeneurship, and about Markov Random Fields. On the wild side: an IP lawyer said in an interview that our company had "the worst name he'd ever seen".
(Privately Held; 51-200 employees; Internet industry)
December 1998 — December 2000 (2 years 1 month)
At Idealab I focused on the problem of optimizing ad monetization. I helped build two different ad serving systems with decision theoretic components. Success was mixed; some of the mathematics was quite good but my inexperience with scalable software solutions led to significant problems.
(Internet industry)
1998 — 1998 (less than a year)
Worked on optimizing ad serving technology. Resulted in a patent (ultimately acquired by Overture.com):
Method and system for optimum placement of advertisements on a webpage: http://www.patentstorm.us/patents/7100111.html
Cognitive Science 1994 — 1998
* Analysis of Direction Selectivity Arising From Recurrent Cortical Interactions: http://neco.mitpress.org/cgi/content/abstract/10/2/353
* A Monte Carlo EM Approach for Partially Observable Diffusion Processes: Theory and Applications to Neural Networks: http://neco.mitpress.org/cgi/content/abstract/14/7/1507
* Robust Sensor Fusion: Analysis and Application to Audio Visual Speech Recognition: http://www.springerlink.com/content/pxut5255633378g2/
B.S , Physics , 1990 — 1994
Erlang, large scale soft real-time systems, decision systems, machine learning, auction theory, functional programming
Yahoo Individual Superstar Award 2005
US patent #7100111 (Method and system for optimum placement of advertisements on a webpage)
National Science Foundation Graduate Fellowship