San Jose, California, United States
Mineset, a resurrected SGI application with tightly integrated and interactive machine learning and visualization.
Architected software to decompose multiple (Word) documents and reassemble them based on corporate best practices, target customer and sales metrics using a genetic algorithm. The assembly process was capable of hierarchical substitution of document parts into the assembled document as well as providing team collaboration to review and sign off on the final documents and to customize each document for the intended recipient based on their role in their company. The database design included distributed hash based objects, natural deduplication and collision free replication. Implemented client word plugin and underlying algorithms as well as integration with Evernote. For the initial vertical, RFP response, the system was capable of automatically assembling the best answers to questions found in the RFP.