Expect Labs, which is behind the MindMeld consumer application, has announced a new developer API. The new API brings the company’s “anticipatory computing” platform to third parties in the cloud.
Use of the API is free. Expect Labs is trying to get it “out there” and see how developers deploy the company’s technology on their applications.
As a practical matter what Expect Labs is offering could be likened to “Google Now in a Box.” Developers and publishers will thus be able offer predictive or contextual search and more sophisticated recommendations and content discovery to users.
Designed primarily for mobile, the technology provides intelligent assistant capabilities to any third party app. It factors in a range of signals including location, time of day, in-app content usage, speech and keyword inputs as well as other information to make content recommendations. All the content processed is the developer or publisher’s own; Expect Labs is not bringing in any outside content.
The API indexes all the publisher’s data as part of the process. By default it can then be used to power site search and for traditional keyword-based queries. But Expect Labs CEO Tim Tuttle says that where the technology really shines is by making contextual and predictive recommendations without manual or spoken keywords.
Here’s the process according to Expect Labs (verbatim copy):
- Point the MindMeld API to your website or database, and it will automatically crawl and index your content and build a custom knowledge graph from your data.
- Drop a few lines of code into your app to begin sending real-time contextual signals from your users to the MindMeld platform.
- Use our versatile API to retrieve search results and recommendations that you can display to your users either proactively or in response to a search query or question.
Expect Labs investors include Samsung and wireless carrier Telefonica. The MindMeld app, which is a technology demo or showcase, offers a form of persistent-search that listens to users’ conversations and offers relevant content and search results without any actual “searching.”
Contributing authors are invited to create content for Search Engine Land and are chosen for their expertise and contribution to the search community. Our contributors work under the oversight of the editorial staff and contributions are checked for quality and relevance to our readers. The opinions they express are their own.
