HelpSocial Releases Platform Version 3 To Help Product Teams Bring AI-Powered Customer Engagement Systems To Market Faster

This month of April marks the 5 year anniversary of our incorporation at HelpSocial. To celebrate the milestone, we would like to announce the general availability of the 3.0 version of our Open API-driven customer engagement platform at HelpSocial!

The new 3.0 updates are centered around two areas: 1) Help businesses accelerate the use of AI throughout their contact centers with less complex development efforts. 2) Provide new tools to help product teams bring customer engagement projects out of development and into the market faster.

Everything we do at HelpSocial is aimed at enabling scalable, cost-reducing service experiences that help businesses improve customer satisfaction scores and grow their number of promoters. We are so proud to announce these 3.0 updates and can’t wait to show them off. 🙂

From brand new tools to application feature enhancements and API additions, there are so many enhancements with this update! Below are some high-level notes about the bigger ones. If you would like to see a demo of these things or get geeky with us on the specific release notes and technical details, contact us here!

 

Open Channel Access

This is a big one! This new set of APIs opens up our platform to handle any type of non-voice message (email, web/mobile chat, etc).

When a message comes into our platform we break the message down into attributes that help a system or agent understand what to do with it. This has helped businesses integrate bots with the social and messaging channels we offer, and help their contact center systems make advanced routing decisions when a customer message arrives. Open Channel Access enables these same capabilities with any non-voice communication channel – whether HelpSocial is the provider of the channel(s) or not.

If you would like to use bots to help respond to an established email or SMS channel, or want to make routing decisions with live chat based on the whole context of the customer situation (vs just the inbound message alone), or just want an easier and faster way to integrate your new WhatsApp channel with your service center, our Open Channel APIs are going to be a great fit.

 

Real-Time Message Enrichment For Conversational AI

When your customer message arrives in the HelpSocial platform a variety of natural language processing and machine learning systems immediately go to work. They seek to provide as much relevant information as possible to answer 2 broad questions: Who am I talking to, and what do I need to know so I can I can help them faster?

In conversational AI, and even with simpler bot programs, understanding the answers to these questions is critical to providing a helpful customer experience. Any product team building these types of systems can attest to how much work is involved in putting all this information together and getting to the usable data points.

Our updates here have done all this work for you – they’re delivered to you with each inbound customer message so you not only understand the customer’s intent in sending the message, but you have the complete language and sentence structure breakdown (verbs, adjectives, punctuation, etc), sentiment scores and emphasis ratings and so much more. This update is aimed squarely at reducing the amount of dev time needed for these projects so you can get to market faster.

Bonus! With our Open Channel APIs, you can now include any non-voice channel you want with your bot program, not just the messaging channels you were looking at before. 😉

 

Rapid Data Modeling For Accelerated Machine Learning

This update opens up our platform for super fast processing of data sets in huge batches. It takes your giant data set, adds all the attributes you want for training your machine learning system, and does it with very little effort on your part.

Machine learning systems are very capable of recognizing patterns but they take a lot of structured data to train. Most companies have to rely on human feedback (corrections), moving at a slow pace of 1 interaction at a time, tell the ML system what is correct or incorrect. For example, if the sentiment of a customer message shows as negative, but it’s actually positive, the agent will need to change the setting from negative to positive. This is effective if the agents are very consistent with making corrections. But it takes a very long time to improve the accuracy rate of the ML system and, if agents are not consistent, it takes even longer (if it improves at all).

Our 3.0 platform updates allow for the upload of thousands to billions of messages to be tagged and associated with attributes, customized to your specific use case. You should still involve agents to give feedback to the ML system, but with Helpsocial-enriched data, the accelerated training of the system skips you years ahead in accuracy and outcome performance.

 

Sound interesting? Schedule a demo with us and get a month to kick the tires for free!