www.CustomChamberMusic.com/calistri-yeh

Fifteen years experience designing and building pragmatic intelligent systems in academic, corporate research, and commercial settings.

Nine years experience managing multi-site software teams and subcontractors.

Proven track record of transforming innovative research into commercial products.

Excellent abilities in design, management, strategic analysis, programming, and communications at all levels.

Interests include data mining, heuristics, information retrieval, intelligent architectures & interfaces, intelligent document management, natural language, plan recognition & planning, statistical & probabilistic reasoning, search, and user modeling.

Senior Director of Context Mining and Content Management (2008-Present)
Director of Context Mining (2006-2008)

Director of Research (2000-2006)
Responsible for all science/research decisions and overall science direction of the company. Provide technical management and leadership for a geographically distributed team of up to 11 scientists, linguists, and analysts. Involved in all aspects of project planning, proposal writing, and customer presentations. Major projects include:

• Semantic Signatures contextual advertising product places highly relevant ads on any webpage in real time. Principal inventor of the underlying Trainable Semantic Vector technology.
• CINDOR (Conceptual Interlingua for Document Retrieval) cross-language search.
• ContentTailor summarization and readable text compression for wireless devices.

Head of Advanced Technology (1998-2000)
Responsible for developing new technology to support the next generation of company products and services. Managed a distributed team of 4 scientists and made direct contributions to projects. Our core research result was a new method for domain-trainable semantic indexing of text. This research resulted in several short-term commercial spin-offs including subject-based document retrieval, real-time clustering and classification, auto-generated trainable custom thesaurus, and an automatic query expander for a natural-language search engine.

Computer Scientist (1996-1998)
Principal architect and implementer of the processing system of MAPIT and PatentMiner, two commercial natural-language data mining systems that search and analyze millions of patents in real time. Designed and implemented the indexing module, query processing module, visualization tools, and overall system architecture. Built working prototype with live data in 5 days; first paying customer in 6 months. Personally processed over 100GB of patent text (subsequent updates were handled by a team of three full-time people). Designed and implemented distributed hardware configuration for fault-tolerant real-time search (50% cheaper and 300% faster than original).

Manager, Artificial Intelligence Group (1993-1996)
Project manager and principal investigator of a three-year government contract with Cornell University and the University of Iowa to build an intelligent planning architecture using techniques from planning, scheduling, machine learning, and distributed computing. Successfully delivered both innovative research breakthroughs and practical solutions to a specific customer problem in large-scale transportation scheduling.

Computer Scientist (1990-1993)
Created blackboard architecture design for a passive radar system. Designed and implemented new method of archiving security audit messages for secure distributed operating system. Designed and implemented network monitoring system to simulate Navy communications traffic. Implemented Bayesian belief network to diagnose radio faults.

PhD in Computer Science, Brown University, 1990 (advisor: Eugene Charniak).

MS in Computer Science, SUNY Albany, 1987.

BS summa cum laude in Computer Science and Mathematics, SUNY Albany, 1986.
Minors: German, Russian, Music, Statistics.

Member of AAAI, ACL, ACM, IEEE Computer Society, Phi Beta Kappa, Sigma Xi.

Classical cellist, composer, and arranger. Principal cellist of Livingston Symphony Orchestra and Synergismus Duo. Founder and CEO of Custom Chamber Music

Author of over 30 journal papers, conference papers, and technical reports. Selected titles:

R.J. Calistri-Yeh, A.M. Segre, and D. Sturgill, "The Peaks and Valleys of ALPS: An Adaptive Learning and Planning System for Transportation Scheduling." In Proc. Int'l Conference on Artificial Intelligence Planning Systems (AIPS-96), Advanced Technology Supplement, to appear 1996.

R.J. Calistri-Yeh, "Applying Blackboard Techniques to Real-Time Signal Processing and Multimedia Network Management." In Proc. 7th Int'l Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems (IEA/AIE-94), 593-599, Austin, TX, June 1994.

R.J. Calistri-Yeh, "Iterative Strengthening: An Algorithm for Generating Anytime Optimal Plans." In Proc. 6th Int'l Conference on Tools with Artificial Intelligence (TAI-94), 728-731, New Orleans, LA, November 1994.

R.J. Calistri-Yeh, "An A* Approach to Robust Plan Recognition for Intelligent Interfaces." In N.G. Bourbakis, ed., Applications of Learning and Planning Methods, vol. 26 of World Scientific Series in Computer Science, 227-251, Teaneck, NJ, 1991.

R.J. Calistri-Yeh, "Utilizing User Models to Handle Ambiguity in Robust Plan Recognition." In User Modeling and User-Adapted Interaction: An International Journal, 1(4):289-322, 1991.

R.J. Calistri, "Classifying and Detecting Plan-Based Misconceptions for Robust Plan Recognition." Ph.D. dissertation, Department of Computer Science, Brown University, Providence, RI, May 1990.