Stephen Keller Email and Phone Number

Sr. Technical Program Manager for Alexa Ads. I helped build, launch, and scale the Alexa Home Screen (AHS) Display Ads program along with Interactive Audio Ads (IAA) product, which launched initially with Amazon Music Free and later with iHeart Radio

Sr. TPM for a number of initiatives in Alexa's Natural Understanding organization. Worked directly with domain customers to do the modeling work necessary to release and improve new Alexa features. Helped design and ship the "Turbo" tool used to make hotfixes to Alexa's NLU models.

Drive the development and release of core speech recognition technology on the machine learning team and sciences team for Alexa. Work with research scientists to produce improved deep neural networks and other statistical models to power the speech recognition that makes Alexa work.

Product manager for the use of voice technology to build new servicing products for American Express card members. Focused on finding innovative ways to leverage state-of-the-art speech recognition in order to build out entirely new capabilities. Areas of emphasis include personalizing the service experience for card members, integrating the user experience across multiple service channels (e.g. mobile, web, phone, and social), and delivering amazing user experiences through natural language.Responsibilities include the creation of new product ideas, getting alignment across business units and with senior management, performing user testing of prototypes, overseeing technical delivery teams, and measuring the success of products post-deployment. Create and implement reporting and A/B testing to ensure products are meeting Key Performance Indicators.

Design and oversee the implementation of custom speech recognition applications. This role involves working closely with clients to perform complete business, functional, and technical analysis of a solution's requirements and then working with our internal development team to build a solutions document that acts as a complete specification for the system to be delivered.Serve as a general speech domain expert and voice user interface (VUI) analyst, guiding clients who may have little experience with speech recognition to build optimal dialogs to accomplish their needs. Light development work, including writing VoiceXML (VXML), and using web technologies such as JavaScript, PHP, SQL, REST/SOAP, etc. Experience with a wide variety of voice platforms including Avaya Experience Portal, Voxeo Prophecy, Genesys Voice Platform, Cisco Voice Portal, Asterisk, and many others.Other duties include: architect of the LumenVox Developer Network (developer.lumenvox.com); writing and managing online knowledge base; high-level technical support for users of the LumenVox software; and serving as subject matter expert for marketing and product management teams.

Agile project manager for suite of speech recognition software. Created milestones and tracked development efforts to ensure regular releases.

Managed documentation library and handled support for speech recognition software suite.

General assignment reporter for a chain of weekly newspapers in northern San Diego. Covered the cities of Encinitas and Carlsbad, including all municipal government activity in addition to general news and local human interest feature stories.Provided editing and page layout duties as needed.

Created, edited, and maintained a wide variety of documentation for internal IT group at the San Onofre Nuclear Generating Station for Southern California Edison. Documentation included software manuals, design and system specification documents, and technical information intended for wider audience, e.g. technical information on the company's intranet sites. Performed light web design and development in HTML, CSS, and ASP.