The ability to use statistical models to determine the most likely outcome to a problem has been the goal of many technology groups over the last few decades. With an eye to Software as a Service (SaaS) offerings, some of these models have been published for consumption using cloud-based APIs.
Broadly speaking these services can be categorized as follows:
Within each category there are multiple options and capabilities developers can use to integrate the AI capabilities into solving today’s business problems.
Through tagging and categorization, the ability to recognize the contents of images has proven to be one of the more fun aspects of AI based Cognitive Services. Questions like “what is in this picture?” or “What are the numbers on this bank check?” have been used as illustrations of the capability in a consumable interactive manner.
The Vision APIs go much further than this and PSC has created applications which have the ability for the end user to create and tag their own set of images for recognition of specific business objects. PSC has worked with our clients to be able to create engaging user experiences where a phone app can be used to recognize text from a vehicle (e.g license plate , VIN, mileage) and using that data create an interactive experience around the vehicle and it’s metrics.
The ability to programmatically translate text from one language to another has been a thing of Science Fiction for years. The Language APIs though, provide a much more granular level of ability to manipulate, understand and facilitate decision making using AI.
PSC has created applications for clients where knowledge based capabilities are able to be translated on the fly into multiple different languages and the review and writing process is significantly reduced for the business.
PSC has also created chatbot applications whereby the bots are able to determine the “intent” of the user’s questions and ensure that their experience using the bot brings joy and not frustration.
The ability to turn text into accurate audio and distill text from an audio file have become seamless in the modern business workspace. Many aspects of the out of the box Microsoft Office experience uses Speech APIs as an integral part of the user experience. In Outlook users can have an email “read” to them. In OneNote and other products users can dictate their text and have it automatically displayed as text within the application.
The Speech APIs enable PSC to take these familiar experiences and extend them into solving more complex business problems.
Cognitive search brings together the ability to utilize the Natural Language Processing, Image processing and text analysis capabilities into one seamless capability to enable search within complex unstructured business data. With the inbuild capability to extract information from images, PDF files and databases the Cognitive Search capability allows for the creation of applications capable of PII detection, sentiment analysis and many other critical data analysis needs.