Using IDOL OnDemand for text analysis in Dynamics CRM - part 1

by Lucas Alexander

Inside any organization's Microsoft Dynamics CRM system, there is a wealth of raw data that can be turned into actionable intelligence if it can be effectively analyzed. This data can be found in emails, case notes, attachments and other records that are created and stored in the course of day-to-day business. In this three-part series, I will show how you can unlock this data using HP IDOL OnDemand's powerful text analysis capabilities. Specifically I will show how you can perform sentiment analysis on incoming emails and also how you can index Dynamics CRM records so that agents can easily see similar, relevant records. Before we get into the technical details, let's take a closer look at IDOL OnDemand.

What is HP IDOL OnDemand?

IDOL OnDemand is a collection of web service APIs that expose a subset of the functionality offered by HP Autonomy. These APIs allow developers to write code that can "understand and engage with human information, including text, images, social media, and app data." Currently IDOL OnDemand offers more than 20 different APIs including sentiment analysis, image recognition, face detection and text indexing/searching. A full list of IDOL OnDemand APIs is available here: https://www.idolondemand.com/developer/apis. Despite the breadth of functionality available in IDOL OnDemand, there are definite differences between it and the on-premise HP IDOL version. You can read more about those differences here.

How do we connect Dynamics CRM and IDOL OnDemand?

Because all IDOL OnDemand APIs are RESTful web services that return responses as JSON-formatted messages, custom workflow activities or plug-ins can communicate with IDOL OnDemand via the System.Net.HttpWebRequest class, but first you need to sign up for an IDOL OnDemand developer account at https://www.idolondemand.com/signup.html. Once you sign up for a developer account, you'll be given an API key that grants access to run IDOL OnDemand APIs. I'll be showing detailed code examples in parts two and three of this series, but in the meantime I suggest you go ahead and take a look at the IDOL OnDemand getting started documentation.

That's all for today. In my next post, I'll show how to analyze inbound emails for customer sentiment. See you then!

A version of this post was originally published on the HP Enterprise Services Application Services blog.