Understanding Knowledge Engine

featureOS Knowledge Engine (KE) is the most helpful knowledge base feature you have ever seen. Think of featureOS Knowledge Engine (KE) as your product’s own Google Search Generative Experience or Perplexity AI, which is trained on top of your knowledge base.

How featureOS Knowledge Engine Works

featureOS Knowledge Engine (KE) is a powerful tool that helps you build engaging product knowledge experiences for your users. It works by indexing your knowledge base and using advanced natural language processing techniques to understand the context of a user’s query.

When a user asks a question, featureOS Knowledge Engine (KE) uses its understanding of the context to provide relevant and accurate answers. It can handle complex queries, understand product terms, and provide answers in multiple languages.

There are three major components to featureOS Knowledge Engine (KE):

1. Indexing and Understanding: Indexing your existing knowledge base articles, the attachments like screenshots, and other information to understand what your product is all about.
2. Searching and Framing: When a user asks a question, featureOS Knowledge Engine (KE) searches through the indexed knowledge base to find the most relevant information and frames the answer in a way that is easy for the user to understand.
3. Learning: featureOS Knowledge Engine (KE) is constantly learning from user interactions and feedback to improve its understanding of your product and provide better answers over time.

How featureOS Knowledge Engine protects your data

featureOS currently (June 2024) uses Google’s Gemini API to provide the knowledge generative experience. Before sending any data to Google’s Gemini APIs, we make sure we process the data to obfuscate any sensitive information that your users or knowledge base might provide. This way, you know that none of the sensitive information is being used by anyone to train the model.

We are about to announce a new enterprise product for which we are building a whole ecosystem of cloud services that we are calling PIE. PIE is enterprise-ready privacy shield that’s build from ground-up which acts as the middleware between your data and other generative AI services.

Why this is pivotal for your product

Users today (including my mom) expect instant answers, and they do not encourage the workflow of finding information. The age of seeking information is coming to an end, and the age of presenting information with context is on the rise.

featureOS KE presents information to the user the way they are used to, and since the UI encourages minimal interaction, the user gets the information they need without having to go through a series of clicks or types or taps.

With featureOS KE, without any manual effort, you can now provide deeper interlinked infromation to your user. For example, if a user asks “How do I integrate with Slack?”, featureOS KE will understand that the Slack integration is available on the Pro plan, and will also provide a note to the user saying “You can upgrade to Pro plan to integrate with Slack”, along with the integration steps.

Where are we heading

We are doing extensive tests to the Knowledge Engine to make sure you get the smoothest experience possible. Our identity is that we ship fast, and we ship often. featureOS KE is no different.

Our dream is to help you integrate one simple search bar, that you can place anywhere in your product, and let the users get instant answers about your product right then and there. We’re not so far from that dream, just a few more iterations.

You can already experience the alpha version of featureOS KE on our help center.