Why Salesforce is killing off Einstein Voice Assistant – VentureBeat

Salesforce is shutting down two of its AI-powered voice services — Einstein Voice Assistant and Voice Skills — as it shifts resources toward its newly released Salesforce Anywhere app, as spotted by Voicebot.ai. A company spokesperson told VentureBeat that voice capabilities remain “a priority” for Salesforce, and that the products it’s discontinuing will inform the development of “reimagined” functionality focused on productivity and collaboration.

Einstein Voice Assistant, which launched in beta last year, was a component of Salesforce’s Einstein Voice — an outgrowth of the company’s Einstein technology that enables customers to navigate cloud services hands-free. One of its ostensible advantages over other platforms was its versatility: It was siloed, restricting data pulls to individual users’ accounts, and it could be “taught” to recognize jargon, acronyms, and slang in an organization’s lexicon.

Einstein Voice Assistant was more than a glorified transcriber. Users could update Salesforce records and create tasks using natural language requests, or tap Einstein Voice Assistant to navigate through Einstein Analytics dashboards and surface metrics like open service cases and performance guidance. Plus, thanks to native integration with popular voice assistants like Google Assistant and Alexa, Einstein Voice Assistant could deliver a daily brief of “key priorities” like upcoming calendar appointments and team pipeline updates.

Meanwhile, Einstein Voice Skills (previously Einstein Voice Builder) was a toolkit that supported the creation of voice-powered corporate apps. From a setup page, developers and admins selected customer relationship management actions — such as updating a field, creating a task, or reading out a prediction — and the fields or objects that informed each of those actions. Apps could be purpose-built for the needs of employee roles or teams, and Einstein Voice Skills afforded control over how information (e.g., next steps and follow-up tasks) was read back and the channels and devices on which apps were accessible.

Both Einstein Voice Assistant and Einstein Voice Skills were available on a range of devices, including third-party smart speakers, smart displays, and an enterprise-focused concept device called the Einstein Smart Speaker.

“We believe AI-enhanced conversational interfaces can transform business processes, which is why we are focusing heavily on innovation in this sphere and building in voice across the entire Salesforce system,” Einstein Voice senior director of product Michael Machado said last November. “Voice allows companies to command and draw insight from Salesforce in a natural way, making an already powerful system even more powerful.”

As of today, Salesforce remains a part of the Voice Interoperability Initiative, a program spearheaded by Amazon to ensure voice-enabled products like smart speakers provide users access to interoperable assistants. When Salesforce confirmed its membership in September, the company said its aim was to ensure users could take advantage of the skills and features offered by Einstein Voice Assistant on a range of devices.

Salesforce’s shifting priorities are perhaps an implicit acknowledgement of the voice market’s competitiveness. In February, Microsoft killed all third-party skills for Cortana, its voice platform, as it refocused the technology for enterprise. Amazon provides a rival enterprise voice service in Alexa for Business, which integrates with Concur to share business travel itinerary information, RingCentral for voicemail readings, and other third-party voice services for corporate customers.

Notably, Salesforce’s decision to wind down Einstein Voice Assistant and Voice Skills follows the departure of chief scientist Richard Socher. During his four years at Salesforce, Socher — previously of MetaMind, which was acquired by Salesforce in 2016 — oversaw R&D initiatives as well as the rise of Einstein cloud AI services for use cases like computer vision, natural language modelstranslation, and personalized CRM search results.