🦄 Making great presentations more accessible.
This project aims to enhances multilingual accessibility and discoverability while maintaining the integrity of original content. Detailed transcriptions and keyframes preserve the nuances and technical insights that make each session compelling.
Overview
📖 AWS re:Invent 2025 - Enhance Contact Center AI With Salesforce and AWS (DAT207)
In this video, Assaf Ben-Gur from Salesforce and Marc Rudkowski from Amazon Connect present their integrated contact center solution combining Salesforce channels with Amazon Connect. They address key challenges: disconnected data, fragmented AI implementations, and agent productivity. The demo showcases Salesforce Contact Center with Amazon Connect (SCCAC), featuring unified agent desktops, Agentforce AI agents across multiple channels (WhatsApp, chat, voice), and Amazon Connect's AI agent assist with real-time recommendations from Salesforce knowledge base. They highlight 12 billion AI-optimized interactions, nearly 100 million monthly voice calls, and thousands of joint customers achieving reduced wrap-up time through automated summarization and enhanced supervisor capabilities managing both human and AI agents.
; This article is entirely auto-generated while preserving the original presentation content as much as possible. Please note that there may be typos or inaccuracies.
Main Part
The Contact Center Challenge: Disconnected Systems and Fragmented AI Solutions
All right. Hi everybody and welcome to the Enhanced Contact Center with Salesforce and AWS. Really excited to be here today. My name is Assaf Ben-Gur. I'm the product lead for Salesforce channels and contact centers, and I'm honored to be joined by Marc. Marc Rudkowski, I'm a principal product manager for Amazon Connect. Been with Connect since we launched it back in 2017. We're a lightning talk, so we've got less than 20 minutes. I've got the timer to get going. We've got about 19 minutes left, so we're going to go through this as best we can because we've got a demo to show as well as part of this. So let's go. Sounds like a plan.
Before we start, we want to share with you the forward-looking statement. Don't make any purchase decisions based on non-generally available product. Let's start a little bit with the challenge of contact centers today. When your customers phone in today, they expect your agents to be super knowledgeable about who they are, what type of relationship they have with you. They expect them to be super knowledgeable and experts in solving the issue that they have. Instead, the experience they're getting is being put on hold because your agent is trying to stitch the customer profile in real time behind the scenes. And when they're ready to action and resolve the issue that they called about, they need to swivel chair between multiple systems.
We keep hearing three things from contact center leaders today. They want to boost agent productivity with high customer satisfaction. They want to reduce operating costs because it's still considered a heavy cost center for the company. And now more than ever, they want to actively participate in growing the top line for the company. They want an active role in growing the revenue for the company. So after resolving the customer issue, they want to free up their agents to quickly focus on upselling and cross-selling.
And AI, specifically generative and agentic experiences, is exactly the technology that can help you transform your contact centers. It's intuitive, it's smarter, it's faster, it's everything we need to take the contact centers to the next level. But it's not easy. We know a lot of those projects just fail. You can't just take AI and just slap it on your data and on your channels and hope for good. Remember that agent that goes and stitches your customer profile because it's all over? Your data is all over, your channels are completely disconnected, your call center is completely disconnected from your digital channels.
Your agents can do either telephony or they can do email or messaging. They're not versatile. They're working off of totally different systems and totally different desktops. And last but not least, AI is the new toy. Everybody in your company wants to do something with AI. Your marketing team is doing something completely different from your sales team that is doing something different from your servicing team. And your R&D and product team, just last week they had this amazing hackathon and they came up with an amazing way using AI to solve a specific use case. That single use case, amazing. But the contact center leaders and the CIO specifically are being left with a lot of point solutions, completely fragmented across.
Unified Customer Experience: Salesforce and Amazon Connect Integration with Agentforce
And this is exactly where Salesforce and Amazon Connect came together to bring you an experience taking channels natively from Salesforce, or bring them over from Amazon Connect and give your agents and your contact center experience one unified, cohesive, standardized way to manage every customer interaction in one single space. From agentic self-service portals to channels like WhatsApp, chat, email, voice that is coming in from Amazon Connect, Apple Messages for Business, Line for Japan, we have all the channels you need in order to standardize all this experience throughout the customer lifecycle journey.
And all those channels are Agentforce, AI agent ready. The same Agentforce builder helps you to build this logic in one place and then distribute it and attach it to any channel that you need. You need a channel to completely automate your email interaction, just use the same agent. You need that channel to work off of your WhatsApp experience, your Line, your web chat, that same agent. And now as of October, thanks to Marc and Amazon Connect, we have a fully integrated solution with Agentforce running on top of the Amazon Connect voice channel.
And our other channels as well, right? And so this is where the key aspect of this, it gives you that optionality between our platforms, right, because we both bring rich capabilities on your digital engagement and your voice interactions. Basically because of that we can run whether it's using Salesforce's native Agentforce agents, or it's using Amazon Connect's AI agents, you have that flexibility in terms of enabling that and having the right fit for each of your use cases in the contact center and CRM. Exactly. And now that you free up your agents from focusing on tier one type of interactions, they have a lot more bandwidth to focus on more value interactions.
Really focus on the interactions where you need a human touch. You need them for dealing with escalation, those more complex types of experiences, and this is where all the assistive AI capabilities come in, things like service replies attached to your email channel grounded by your knowledge. That same email reply application and service reply application apply to your messaging channels, your WhatsApp, your web chat, your Apple Business for messaging. At the end of the call, you want to reduce that wrap-up time. Call summarization helps you to reduce minutes out of it so you can free up your human agent to focus on the next interaction.
And last but not least is your supervisor. It's a whole new world for your supervisors. Yesterday they just managed the human labor workforce. Now they need to manage all those AI agents. This is why we transformed our Omni Supervisor to be a command center for service. Really bring in a cockpit, one single unified workplace that allows you to manage both your human agents and your AI agents, so your supervisor can get alerts in real time and red flags raised based on conversations that your AI agents are driving and then be able to listen in and interject where needed.
Salesforce Contact Center with Amazon Connect (SCCAC): New Capabilities and Roadmap
But there's so much more. I know, Mark, you guys have been busy in the last nine months building the next wave of innovations. Yeah, so let me talk a little bit in terms of what we've actually announced this week, and I'll talk a little bit in terms of the integration that we have with the Salesforce Contact Center with Amazon Connect. Yeah, thank you for the clicker.
Back in March, we announced and launched Salesforce Contact Center with Amazon Connect. I know that's a mouthful, so for the rest of the presentation, you're going to hear me use the term SCCAC because it's a lot easier to say from that perspective. Back in March, when we announced and launched this, essentially what it was, was we worked super closely with Salesforce and we saw a lot of success that we had with Agentforce Voice in terms of enabling that with a lot of our core contact centers. But what customers really wanted was they wanted additional channels. They wanted additional Connect capabilities that were unified as part of Agentforce Voice.
So when we released this in March, we were actually the first contact center to implement Bring Your Own Chat, which essentially takes our native chat and it integrates with Salesforce's chat capabilities. So the advantage of that is that when you're in Connect and you're in our IVR experience, you can leverage our voice, you can leverage our native chat, but when it gets to the point of actually escalating to an agent, we actually then send it in through Salesforce chat UX and it essentially pops up as a native Salesforce chat within Agentforce Voice. So the advantage of that is from an agent perspective, whether they're using Salesforce chat or our native chat, it looks the same. You get that deep, rich integration across the capabilities.
So for example, being able to have those preconfigured replies, right, to be able to answer those customer inquiries that are often happening and giving agents shortcuts to quickly answer questions. We can enable that really easily and natively within Salesforce as part of that. The other aspect of what we announced this week is now we started to really then start to expand additional Connect capabilities. So what you probably heard a lot this week with Connect is us announcing our native AI agents integrated as part of Connect, and part of that is really giving you out-of-the-box AI agents that you can start to leverage both from a self-service perspective as well as an agent assist perspective.
So what we announced this week is essentially the next version of SCCAC which now allows you to add agent assist capabilities with Connect AI agents natively within Agentforce Voice. So the advantage of that now is essentially from a self-service perspective, whether you're using our voice or our chat channels, you can have that rich AI agent experience and then handle all of that interaction through the Agentforce Voice experience. And then as the agent is trying to continue to resolve the issue with that context, for example, they could be plugging into Salesforce knowledge base for getting all the SOPs and recommendations. And essentially as they go through that, and I'll show a demo of this in a little bit, have that interaction, walk through that and then give the agent essentially the right recommendations to solve that issue as quickly as possible.
Once the chat or the voice call is over, at that point, give them the summary of the interaction that they have so they can quickly update the Salesforce case to ensure that, as Assaf mentioned earlier, saving often minutes from the agent's time in terms of improving their overall productivity to not only solve the issue, but then update the issue from an after-call work perspective. And what you see here is you see we've got quite a rich roadmap in terms of capabilities that we've recently announced in terms of overall sets of features. What I just talked about here was essentially our agent assist and summarization capabilities. I'll do a demo of that in a few minutes.
What's that? Yeah, okay, yeah, so, and then real-time transcription for additional languages. So back at Dreamforce, Salesforce announced the additional support. So that runs a lot on our core underlying capabilities like Amazon Transcribe. And it really talks to the global capabilities that we have as a better together story.
So Connect brings that, particularly around our telephony capabilities. It's global telephony. We bring that at scale in terms of managed telephony across the entire world, and part of that is really now with Salesforce. From an agent experience, you get all the language support around all of that as well. So again, we are truly a global solution in terms of all the capabilities that you get between Salesforce and Amazon Connect.
We had also announced Agentforce Voice with Amazon Connect. So in terms of supporting third-party voice experiences, we now enable that as part of Amazon Connect as well. And then finally, from a workforce management perspective, if you're using our chat and our voice capabilities, you can leverage our workforce management or FCS capabilities, forecasting capacity planning, and scheduling capabilities natively within Connect. And from a roadmap perspective, we're working with Salesforce to start to support their channels in the future as well. So again, you can see we continue to evolve the overall capabilities and integration between the two in terms of overall feature sets and just richer and richer integrations overall.
Live Demo and Customer Success: AI-Powered Agent Assistance in Action
So let's jump into a demo. Let's see, I'm looking at my clock. I got 8 minutes and 34 seconds, so plenty of time here. So let's see if we can go through the demo. All right, perfect. And so this is a bit of a click-through demo because nobody trusts me doing live demos, especially on stage when I've only got 20 minutes to walk through that, actually less than that, 8 minutes now.
So we've got, just to set up some background, this is Any Company, it's an energy company use case that we talked through here. We've got Nikki as the customer. She's coming in. She has a problem with her generator, and we're going to help her basically solve that issue within a few clicks overall from that perspective.
So the chat comes in. Essentially what you see here is, actually, this is always the challenge here. There we go. And so we have the chat here with Nikki, and she'd say, thank you for contacting us. Now again, this is for the sake of speed, we're doing a fairly simple use case here, but you can imagine this could be using one of our AI agents. It could be using Agentforce agents in terms of creating that interaction, that self-service experience in terms of overall capabilities around that. Once we escalate to an agent, which is what we're going to show here, essentially all the context of that interaction gets handed over to Nikki as part of that.
So essentially at this point, the chat comes in. Nikki gets this essentially omni item in terms of being able to answer the chat as part of that. This is her dashboard, by the way, so she has a rich dashboard in terms of seeing how she's doing overall across her day. Right, we click on that, and now at this point you see the chat, the messaging window. Again, you can see it's, if you're familiar with Salesforce's chat window, it looks very similar, it's identical to that in terms of that overall interaction. They're having that interaction from that perspective.
Now a few clicks later, she's asking essentially, I need guidance on a generator installation, can you please help? Sure, right? So at this point, in real time, the Connect AI agent, which is also hooked up to a Salesforce knowledge base, is now understanding the intent and the ask that's happening as part of the chat. And as part of that, it's now going to provide a recommendation back to the agent, where the agent can then quickly read that and then provide that guidance back to Nikki in terms of solving this overall generator issue.
In addition to that, it's also going to provide the links to the sources. So if the agent wants to, again with a human in the middle, being able to look at in terms of the sources of that data and do a little bit deeper research if needed, they can understand where that information's coming from, which is great. They can also search as well. So if it doesn't quite meet in terms of what they're looking for in terms of a recommendation, they can easily there and continue to go search through the knowledge base for additional answers as part of that.
So great, the agent responds as part of that interaction, and then we solve Nikki's problem. And then at the end, you see it there up in the upper right corner is the Amazon Connect summarization. So now we've taken that end-to-end interaction, that half from a chat perspective, and we've summarized that, and then essentially we can update the case within a few clicks, and we've solved the overall issue again. So you can see here that we've taken the native Amazon Connect capabilities and we've integrated into the Salesforce Agentforce Voice experience and really providing that overall unified experience overall.
So Marc, the agents don't need to log in into multiple systems. The agents just work off of that one single unified desktop. Exactly, exactly. So we've removed a lot of the heavy lifting, right, that our customers will have to do in terms of integrating these capabilities, and we've made it from an out-of-the-box perspective, simple, easy to set up, configure, and start to enable. Perfect.
All right, so let's switch back through the presentation. All right. So some of the things here to note here too is Amazon Connect, we have tens of thousands of customers supporting more than 160 million interactions a day.
This really speaks to the scale at which Amazon Connect is leveraged and deployed. You can see here the numerous customers, and we have a lot of joint customers that are also Salesforce customers that we work super closely with. Vista, Truist, and Air Canada are all great examples in terms of who use this integration today and have continued to scale this overall across solving their key contact center capabilities and problems. We've seen a lot of success from that overall.
Just as a reminder, oh, go ahead, did you want to say something? We have thousands of joint customers already, thousands of joint customers growing really nicely. We're seeing double digit growth, almost hitting 100 million voice calls a month, and this is an amazing number for Amazon Connect across all customers. We are the largest contact center in terms of leveraging this with Salesforce in terms of the integration capabilities that we have.
For those of you that were in the session yesterday, I brought this up as another reminder. Last year, the number of customer interactions that are using AI within Amazon Connect to optimize those interactions was about 6 billion. We've doubled that now to 12 billion. This continues to show the fact that we have used AI from the beginning with Connect, and we continue to use that journey and enable that natively within Connect. We're also working closely with our partners like Salesforce to give you those capabilities as well from a joint perspective, a better together story as we work together to continue to evolve what I'd say is a very fast, rapidly changing space from that perspective.
It is the number one use case, the contact center with AI. We're here working together in terms of how do we best solve your use cases as end customers and large enterprises to make sure that we're meeting that to the speed and capabilities that you need from an overall contact center and CRM perspective. With that, I've got 2 minutes and 40 seconds, so I think I did all right.
What I'll say is we've got our booths here, so definitely come visit the Salesforce booth, booth number 739. I think it's right over there from that perspective. Then come visit the Amazon Connect booth. That's hard to miss. It's not a big idea, but the AWS one that's right in the middle. I've got a QR code here as well. Obviously, for those of you who have your account teams, feel free to reach out to them. They can definitely get you up to speed on all the capabilities.
Follow me on LinkedIn, follow us on LinkedIn. I posted a bunch of stuff up there as well, so you can take a look in terms of links in terms of how to get more information about Agentforce Voice or SCCAC. We've got all that information up there, and obviously we're here as well, so happy to have the conversation and talk some more about it. Regardless of what footprint they have, whether they have other point solutions that they want to go and transform and take it to the next stage potentially with Salesforce and Amazon Connect, or whether you already invested in Amazon Connect channels or you want to evaluate Salesforce channels and bring it all along, we have the solution for you.
We've built it in a modular way that can bring all this modularity and all these components together meaningfully for your agent and supervisor. We're looking forward to any next steps here. With that, we thank you everybody for attending today, and I did it with a minute left. Thanks everybody. Have a good rest of your day.
; This article is entirely auto-generated using Amazon Bedrock.
Top comments (0)