Providing a great customer service experience is tough. To do so, you have to meet your customers where they are and get them to engage, understand their questions accurately, and then solve their problems efficiently. As we’ve discussed in other posts, Watson Assistant does a great job helping you produce this kind of experience.

But what happens when you need to expand your assistant to support more of your customers’ needs? A key concern for any of our enterprise clients in expansion mode is whether the assistant can speak all of the languages that their customers speak. They want confidence that once they start building in the first or second language, that the full set of languages will ultimately be supported as well.

The good news here is that Watson Assistant now supports every global language, but before we jump into that, let’s take a look at how our clients tried to get additional language support before today.

Making a square peg fit in a round hole

Up until now, Watson Assistant supported 13 major different languages natively within the product (you can see the full list here). This means that nordic languages, Russian, Thai, Greek, and others spoken in certain parts of Asia and Africa weren’t officially supported. If clients wanted to create a new assistant to handle an unsupported language, they would have two options:

1. Purchase and use a language translation service in front of their existing assistant
2. Try to find a supported language that was close to the language they wanted (like using a Czech language model for Russian)

Using a language translation service is prone to the effect of compounding errors. If anything goes wrong in the original translation, the language model being used is likely to exacerbate those errors and prevent you from correctly understanding users. And if you decided to go with a similar language model — like using the Czech language model for Russian text — you’re likely to run into issues with the underlying grammatical structure of the language. In essence, each language model is performing a preprocessing like tokenization specific to the supported language. Thus, if you tried to feed in an unsupported language, you might run into major issues and, thus, incorrectly understand users.

Introducing the Universal Language Model

Watson Assistant is upending this paradigm and introducing a Universal Language Model to help! This new feature builds on the breakthroughs in intent classification to create a language model that can be used in over 100 different global languages. Using advanced natural language processing and AutoML, Dialog and Actions skills in Watson Assistant can adapt and train to a previously unsupported language. Now, Watson Assistant clients can build assistants in each and every language that they want to support, taking advantage of the best-in-class machine and deep learning techniques to solve end-user issues anywhere in the world. If you compare this to the 10 to 30 languages most other vendors support, it’s clear Watson Assistant is leading the field.

How can you use this new Universal Language Model to expand support to new languages? In Watson Assistant, you can create a new Dialog or Actions skill and select the Universal Language Model from our drop-down list. Once you do, your assistant will be ready to understand users in a new language outside of the 13 languages we support natively today!

Check out the Universal Language Model in Watson Assistant today! Happy building!

And if you’re new, you can learn more and sign up for Watson Assistant here.

An assistant for any language was originally published in IBM watsonx Assistant on Medium, where people are continuing the conversation by highlighting and responding to this story.

“Without continual growth and progress, such words as improvement, achievement, and success have no meaning” — Benjamin Franklin

Watson Assistant has always been a pioneer in the chatbot market (though we prefer to call them assistants). We’ve continued to evolve our product over the years from what started as a stateless developer API (see below) — with all of the frustrations and challenges of building an external application around that — into a holistic customer care product designed to solve customer and employee challenges — the first time.

2020 was a year like no other. In particular, customer and employee services were challenged in unimaginable ways with lockdowns going into effect, remote work becoming the norm, and historic call center volumes pushing agents to their limits. Let’s take a look at how Watson Assistant has evolved to meet the daunting market challenges that await for 2021 and take a look at our updated pricing.

1. Getting to first contact resolution

As contact centers and help desks continue to deal with historic call and chat volume, it should come as no surprise that resolving questions quickly and successfully — the first time — is one of our biggest priorities as a product. When it comes to solving the problem of poor first contact resolution, it comes down to two critical challenges: getting users to engage and successfully understanding their question.

When it comes to engagement, here’s how Watson Assistant can become a tool your customers actually want to use:

1. Phone, SMS, and WhatsApp integrations: Watson Assistant now supports phone, SMS, and WhatsApp natively through our product. And with the phone integration, clients can quickly deploy a voice assistant as a modern replacement to a legacy IVR (through their existing telephony provider)! Plus, our speech services (which are automatically provisioned when you create a voice integration) come with natural sounding voices that we created through the use of advanced AI.
2. The best intent recognition model: If you’re going to understand your customers without using hundreds or even thousands of training examples you better have some good AI. Watson Assistant was recently shown to provide the best intent recognition accuracy among major competitors including open source products. That means Watson Assistant will help you understand your customer's questions without needing them to always “try phrasing it in another way” as with some chatbot vendors
3. Home Screen & the webchat client: We previously released a webchat client to help our clients quickly deploy an assistant to their website. But it’s not always easy to understand what a virtual assistant can help with. That’s why we introduced Home Screen to go along with the webchat client. Home Screen allows clients to display a welcome message along with common intents the assistant can help with to guide users. In internal testing, we found that Home Screen drove a 50% increase in engagement on the first interaction.
4. Even better automation to get to the right question: Even with great AI, sometimes users can ask questions in a convoluted way. Watson Assistant has always had disambiguation to help guide users to the right place. But we’ve now added even more enhancement with Suggestions. Suggestions uses AI to understand when a user is stuck and proactively offer suggestions for intent or questions they may be interested in

Put together, all of these new features (plus ones that have been around for a while like disambiguation and the webchat client) help ensure that Watson Assistant is an engaging experience your customers actually want to use with the AI to ensure you’re able to understand them — the first time.

2. Solving the fragmented customer care tech stack

Businesses and organizations are and should be investing in best-of-breed tools that work for their business. They shouldn’t be forced to choose applications or tools because they work with their chatbot. Watson Assistant has been designed as an open ecosystem so that our clients can connect to the tools, systems, and applications they’ve already invested in.

But in addition to this open ecosystem, clients increasingly need the ability to orchestrate the end-to-end experience. If you’re using a bot platform as just a developer service/middleware, this leaves you on the hook for creating a big, complicated orchestration app: mastering all of the channels/systems you need to connect to, managing the state of the conversation, and much more. Now, Watson Assistant has evolved to handle way more of this for you!

1. Pre-event webhooks: Our new pre-event webhooks are a core part of our ability to orchestrate the entire user experience. With pre-event webhooks, you can now manipulate information before it hits Watson Assistant. This is critical for use cases like removing PII data, user authentication, and language translation.
2. Post-event webhooks: We’ve also added post-event webhooks, which means you can now orchestrate systems or modify responses just before they reach the end user on a channel.
3. Logging webhooks: All of your conversational data can now be sent as events via a logging webhook — this allows you to more easily use and analyze your data in other tools or warehouses (in real time).
4. Content crawlers: The search skill provides responses from your existing content sources, and can crawl all sorts of data sources like web pages, PDFs, Sharepoint, Box, Salesforce, etc…
5. Escalate to agents via service desk integrations: In addition to our out-of-the-box service desk integrations with Salesforce and Zendesk, we’ve expanded to help our clients connect to other leading customer care platforms for live agent hand-off. With bring your own service desk, we offer pre-built code packages to help integrate to Twilio Flex, Genesys Cloud, and NICE inContact. Plus, we offer base code to help make integrations to other customer care platforms.

To learn more about all of our orchestration capabilities, head to this post.

3. Scaling beyond simplistic FAQ bots

While solving the problems of first contact resolution and a fragmented customer care tech stack is critical, clients also need a way to scale their virtual assistant to increasing complex use cases. These use cases typically involve multiple variations that will depend on specific user input.

1. Actions: We’ve recently released Actions in beta. Actions is a dramatically simplified way to build out conversational flows inside of Watson Assistant. Instead of needing to transverse between intents, entities, and dialog; now all of the core functionality has been built into one page. Actions improves on the visual editor available in our dialog builder today and will include additional functionality like the ability to deliver rich media back to users.
2. Coverage & containment metrics: The addition of coverage and containment metrics to our analytics section of Watson Assistant gives clients the ability to better analyze the performance of their assistant. Coverage will signify if you’ve built out the right intents to handle the questions your users are actually asking while containment will give you a view into how often your conversations are getting escalated to an agent.
3. Autolearning: With Autolearning, you can automatically improve user responses over time and create a more efficient and accurate assistant with minimal effort. Autolearning learns from disambiguation clicks. When a user selects a given option in disambiguation, Autolearning uses that input to update the underlying weights in our machine learning model for intent recognition. Over time, this will result is more accurate responses from your assistant back to users.
4. Language expansion: We’re constantly working to improve the number of languages we support with Watson Assistant. We’ve released new support for our intent recommendations feature in French, German, Italian, and more. You can check out all the latest languages currently supported in Watson Assitant here.

Our New Pricing

As we’ve evolved, we’re also updating our pricing to reflect that evolution. The new features we’ve released have collectively reduced the total cost of ownership of Watson Assistant compared to other vendors. Now, clients can build a robust virtual assistant on any channel, seamlessly connect to back-end systems and applications, orchestrate the end-to-end experience, and scale to virtually any use case.

Here are the specifics on what’s changing on March 1st, 2021:

Plus Plan: Our Plus plan will start at $140 per month with 1,000 MAUs included ($0.14 per user). Additional MAUs can be purchased for $14 per 100 MAUs. To get started with the phone integration, it’s a simple add-on charge of $9 per 100 voice MAUs ($0.23 per user with digital & voice combined). And you don’t have to deal with the per-minute speech charges separately — it’s all included in the price!

Enterprise Plan: We’ll also be launching a new Enterprise plan as an evolution of our Premium plan. Enterprise will come with everything in the Plus plan with higher limits, additional enterprise features (like the options for data isolation and HIPPA-readiness), and a more cost-effective price than the Plus plan at higher volumes.

Look out for additional communication and information on March 1st, 2021 as the Enterprise plan and the new pricing goes into effect!

Come check out the new features today. New customers can get started for free with our Lite plan or Plus plan trial! Happy building!

Watson Assistant is Evolving (again!) was originally published in IBM watsonx Assistant on Medium, where people are continuing the conversation by highlighting and responding to this story.

Introducing pre-built integrations with Zendesk and Salesforce for live agent handoff

Have you been forced to call a customer service line recently? If you have, chances are this story may sound all too familiar — hearing a robotic set of numbered menu options (which have somehow always changed) and then continuously pressing 0 until you finally hear the magical words “Okay, I’ll connect you to an agent”. Unfortunately, your reprieve only lasts a few moments until you’re forced into a digital queue doomed to repeat the same story 3 or 4 times before you finally get to the right department. Customer service doesn’t need to be this way. In a world where every minute matters, great customer service isn’t just a differentiator — it’s imperative.

We understand how hard it is to create a customer experience that can delight consumers. Most customer service solutions only solve part of the problem. Your business needs a product that can work right away, be deployed easily, and connects to the systems you’ve already invested in. We get it. That’s why we’ve updated our virtual assistant product, Watson Assistant, to seamlessly connect with Zendesk and Salesforce. (For an overview on Assistant, click here) How can these pre-built integrations help your organization? Let’s take a look.

Connections that Matter

Watson Assistant isn’t your average chatbot. We’ve built it using best-in-class AI technology that’s used to understand what a user is asking. For more information on this, please click here. We know, however, that some problems require a human touch. With our new pre-built service desk integrations, users can quickly connect to a live agent without ever leaving the chat window. But we didn’t stop there. We wanted to make sure that when a user engages with a live agent they are able to have a meaningful interaction. To facilitate this, we automatically summarize the conversation so your agent can begin helping right away. If they need additional context, the agent can also see the entire chat history. This means users don’t need to repeat the same story over and over again. Most importantly, it means your customers get the experience they deserve.

Let’s take a deeper look at these integrations from the user and agent perspectives:

Time to Value

Excellent service. Meaningful connections. You may have the best intentions in designing your customer experience but at the end of the day, providing great customer service can be expensive. Previously, building service desk integrations could take months of development work with considerable cost to a business. Now, those integrations only take minutes to set up with a few simple clicks.

If you want to give your customers an experience they’ll tell their friends about, come build with Watson Assistant. To learn more about setting up these integrations, check out our how-to videos for Zendesk and Salesforce. If you are not yet a Watson Assistant customer but you’re ready to start building, sign up for Watson Assistant here.

Leave No Problem Unresolved — Connect to a Live Agent with Watson Assistant was originally published in IBM watsonx Assistant on Medium, where people are continuing the conversation by highlighting and responding to this story.

Introducing pre-built integrations with Zendesk and Salesforce for live agent handoff

Have you been forced to call a customer service line recently? If you have, chances are this story may sound all too familiar — hearing a robotic set of numbered menu options (which have somehow always changed) and then continuously pressing 0 until you finally hear the magical words “Okay, I’ll connect you to an agent”. Unfortunately, your reprieve only lasts a few moments until you’re forced into a digital queue doomed to repeat the same story 3 or 4 times before you finally get to the right department. Customer service doesn’t need to be this way. In a world where every minute matters, great customer service isn’t just a differentiator — it’s imperative.

We understand how hard it is to create a customer experience that can delight consumers. Most customer service solutions only solve part of the problem. Your business needs a product that can work right away, be deployed easily, and connects to the systems you’ve already invested in. We get it. That’s why we’ve updated our virtual assistant product, Watson Assistant, to seamlessly connect with Zendesk and Salesforce. (For an overview on Assistant, click here) How can these pre-built integrations help your organization? Let’s take a look.

Connections that Matter

Watson Assistant isn’t your average chatbot. We’ve built it using best-in-class AI technology that’s used to understand what a user is asking. For more information on this, please click here. We know, however, that some problems require a human touch. With our new pre-built service desk integrations, users can quickly connect to a live agent without ever leaving the chat window. But we didn’t stop there. We wanted to make sure that when a user engages with a live agent they are able to have a meaningful interaction. To facilitate this, we automatically summarize the conversation so your agent can begin helping right away. If they need additional context, the agent can also see the entire chat history. This means users don’t need to repeat the same story over and over again. Most importantly, it means your customers get the experience they deserve.

Let’s take a deeper look at these integrations from the user and agent perspectives:

Time to Value

Excellent service. Meaningful connections. You may have the best intentions in designing your customer experience but at the end of the day, providing great customer service can be expensive. Previously, building service desk integrations could take months of development work with considerable cost to a business. Now, those integrations only take minutes to set up with a few simple clicks.

If you want to give your customers an experience they’ll tell their friends about, come build with Watson Assistant. To learn more about setting up these integrations, check out our how-to videos for Zendesk and Salesforce. If you are not yet a Watson Assistant customer but you’re ready to start building, sign up for Watson Assistant here.