In order for Harmony to access all the information and transactions happening in SYSPRO, a data mining service was created.

This mining service monitors the SYSPRO database and analyses its transaction log. It cleanses the data and then stores it in a relational database ready to be processed by Harmony and the Rules Based Engine.

	This service can reside anywhere in the SYSPRO infrastructure, however we recommend that it is installed on the same server as the SQL Server.

In addition, data sentiment enables a complete spectrum picture of customer opinion about your services or products for informed and strategic decision making. Visual cues (the equivalent of emoticons), help alert users by attaching sentiment to specific items.

Harmony provides the equivalent of a news feed that's auto-populated by conversations with relevant information, based on key ERP components that you choose to follow.

Instead of needing to go into the system to retrieve pertinent information, the system intelligently pushes it to you.

Similarly, you can opt into following any other key information in SYSPRO (e.g. stock codes, suppliers, sales orders, etc.).

For example: If you choose to follow the customer Bayside Bikes, your newsfeed will update every time there is an activity for that customer (e.g. when the customer's information has changed, or when a sales order has been created, or even when the customer has returned something).

In addition, some intelligent algorithms monitor the content that your peers follow (as well as what is trending in the database) and these insights are surfaced in your newsfeed.

The collaboration aspect of Harmony brings SYSPRO users together via a chat method, similar to that of WhatsApp or Skype. However, unlike traditional chats, Harmony allows users to bring the context of their ERP data into the conversation using social media techniques (e.g. # hashtags).

Harmony lets you post beats (i.e. conversation posts within the Harmony portal) like you would on Facebook or Twitter. However, you have the added benefit of those beats being able to contain data context with the use of @ signs and # tags. For example: If you enter #Customer abc. Balance, Harmony returns that customer's balance immediately and enables you to beat about it.

Therefore, when collaborating, users have full and secure access to the data in their SYSPRO databases. This enables them to query information directly in the chat, such as stock information, sales order statuses, customer information, etc.

Users can reply and like messages that have been posted, thus elevating their importance or relevance.

Additionally, Harmony makes use of cloud-based cognitive services that detect conversation sentiment. It will detect if the conversation's sentiment is positive, negative or neutral, and indicate accordingly.

Most social media applications (like Twitter or LinkedIn) use intelligent machine learning algorithms to detect trends from the messages that are being posted, and then surface that information to the user according to their relevance. In a similar fashion, Harmony uses predictive analytics to reveal trends that provide actionable insight. The differentiator is that not only does Harmony process machine learning and trend detection on conversations, it also processes these on the core SYSPRO database.

Harmony will identify transactions happening on the SYSPRO database, identify the key fields on which the transactions are happening, and then perform trend detection and aggregation on this data. It will then present this information to the user in the form of a chart and reflect what is currently trending in the SYSPRO database.

For example: A stock code could be used in a large number of sales orders, as well as some purchase orders and RMAs. In this case, that stock code would be trending.

This simple piece of functionality brings critical information to the forefront without you having to ask for it or consolidate reports across the different modules.

The rules based engine works together with the data mining service and allows you to create complex triggers and perform actions without using SQL triggers.

Harmony also uses the rules based engine. If a user follows a particular item, a rule is automatically added to the engine and when the rule executes, a beat is posted to the Harmony database.

	Although currently you cannot add, maintain or delete rules, SYSPRO is developing a new interface to enable this functionality.