Introducing SensoryCloud.ai: Part 1 of a Multipart Series– Flexibility

After a quarter century of running embedded or “on the Edge” Sensory is moving into the cloud with the opportunity to offer hybrid solutions with more Flexibility, Accuracy, Features/Technologies, Privacy and Cost advantages than ever before.

This is the first of a multipart blog series to introduce SensoryCloud.ai and the key advantages. This blog will focus on Flexibility.

One of the reasons many customers have turned to Sensory over the years is that we are willing to customize for what our customers need. The approaches by Google and Amazon are fairly cookie cutter based approaches with predetermined ways of doing everything. This works well at their scale and often appeals to smaller companies that don’t have the budgets or need for the very best customized solutions.

The SensoryCloud will offer a lot of Flexibility, so let’s look at some of the ways we can work with our customers:

• API or SDK – We can provide direct access to the SensoryCloud through API for usage-based pricing, or we can provide platform-based ultralight SDKs that enable fast gRPC based communications to the cloud
• On-premise or in Cloud (AWS/Azure /GC) – Sensory provides containerized code that runs on the cluster host our customer selects
• Management/Maintenance – Sensory can remotely handle all deployment and inference server maintenance or pass one or both of these responsibilities on to our customers
• Pricing/Licensing – We are open to usage-based fees, time-based structures, and we can even offer source code buyouts (that gets expensive!)
• Data usage and Modeling – Sensory enables data sharing, data owned strictly by customers, custom model building, automated model building, whether data is collected at all, and where data is stored.
• Hybrid Solutions – Functionality can be completely deployed in cloud, completely on device, or a hybrid mix. For example, wake words can run low power on device and get re-validated with highest performance in cloud. Feature extraction, inference and other functions can be independently deployed either on device or in cloud to optimize for product cost, privacy, power consumption, and needed features/benefits
• Multiple Features – Customers can select from a vast array of technologies and pick and choose those they want and easily access, test and deploy. Functions like Speech to Text, Text to Speech, Custom Assistants, wake words, Computer Vision, Voice Biometrics, Sound ID, and Face Biometrics can all be accessed through the SensoryCloud.