AWS re:Invent 2022 - [NEW] Resolve customer issues faster w/Amazon Connect agent workspace (BIZ210)

AWS re:Invent 2022 - [NEW] Resolve customer issues faster w/Amazon Connect agent workspace (BIZ210)

AWS re:Invent 2022 - [NEW] Resolve customer issues faster w/Amazon Connect agent workspace (BIZ210)

Amazon Connect agent workspace consolidates all your agent features, such as customer information, cases, tasks, knowledge articles, and workforce schedules, into a single easy-to-use user interface. Agent workspace includes Amazon Connect Flows so that contact center supervisors can configure and provide step-by-step guidance to agents. This helps agents resolve issues faster and improves customer satisfaction. Join this session to learn how to create flows that display information an agent needs to take action and resolve an issue quickly.

Learn more about AWS re:Invent at https://go.aws/3ikK4dD.

Subscribe:
More AWS videos http://bit.ly/2O3zS75
More AWS events videos http://bit.ly/316g9t4

ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.

AWS is the world’s most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—are using AWS to lower costs, become more agile, and innovate faster.

#reInvent2022 #AWSreInvent2022 #AWSEvents


Content

0 -> - All right well, good morning everyone.
5.01 -> It's good to see some familiar faces in the audience
7.74 -> and lots of new faces as well.
10.86 -> Welcome to our session today,
13.02 -> resolving customer issues faster
15.78 -> with the Amazon Connect agent workspace.
19.8 -> If you are in this room right now,
21.57 -> you probably operate a contact center.
25.8 -> Maybe you're already using Amazon Connect
29.04 -> or you're considering Amazon Connect
31.56 -> for your contact center.
35.49 -> You are also likely facing some challenges.
39.48 -> Perhaps you maintain software infrastructure
42.69 -> that your agents currently rely on to help customers,
47.58 -> and you're looking for alternatives
50.46 -> to drive agent efficiency.
55.14 -> Well, we're excited to be up here today
57.24 -> and tell you about a new launch
59.19 -> that was announced yesterday
60.81 -> in Adam Selipsky's keynote address
63.39 -> that we think will help solve some of these challenges
67.23 -> with cloud-based technology and intuitive agent interfaces.
81 -> I'm Sandeep Rao, go-to market specialist
84.72 -> with AWS productivity applications.
88.74 -> And this is Jack Hutton,
90.48 -> senior product manager for Amazon Connect.
94.17 -> Let's get started.
96.84 -> All right, so many of you
99.12 -> already know what Amazon Connect is.
101.55 -> For those of you who do not,
103.65 -> Amazon Connect is a true cloud-based contact center
108.09 -> that allows you to scale up to any size workload
111.99 -> and you only pay for what you use.
115.17 -> It's very simple to get started.
117.51 -> Anyone from a non-technical business leader
122.28 -> to experienced contact center veterans
125.01 -> can begin innovating on behalf of their customers
128.79 -> with our self-service graphical user interface.
134.55 -> You can design conversational interactions
137.61 -> that are natural and personal using our speech recognition
142.59 -> and natural language understanding technology,
145.47 -> the same technology that powers Alexa.
149.4 -> And the combination of these benefits
151.83 -> allows organizations to design customer experiences
155.7 -> that meet expectations, drive business goals,
159.3 -> and reduce costs.
163.62 -> - Connect is one of the fastest growing services
166.14 -> in AWS history.
167.76 -> Tens of thousands of customers use Connect
170.28 -> creating over 10 million contacts in total per day.
174.15 -> Large customers such as National Australian Bank,
177.06 -> Adidas and Delta love that they can experiment
180.72 -> and move at a rapid pace like a startup.
183.39 -> Meanwhile, smaller customers such as Answerconnect,
186.6 -> Slice Pizza and Chartspan love that they can get
189.39 -> an enterprise grade solution at startup prices.
193.11 -> Overall, we value our customer's input,
195.36 -> and it's very important that we listen to customer feedback
198.3 -> in order for us to innovate and allow you
200.31 -> to rapidly expand your customer service operations.
203.37 -> And so I just wanna give a quick thank you to all of you
205.53 -> and all our customers for trusting us
207.51 -> as you overall transform your customer experience.
212.73 -> - All right, so you've seen some organizations
214.62 -> who use Amazon Connect today,
216.6 -> but how do agents use Amazon Connect?
220.35 -> Well today agents can use our free out of the box
225.09 -> agent workspace to get work done more efficiently.
229.71 -> This agent workspace is available virtually anywhere
233.85 -> via web and single sign-on.
237.33 -> All of our agent empowerment capabilities
239.34 -> are unified in a single pane of glass,
242.19 -> but again, you only pay for what you use.
245.64 -> The interface is intuitive,
247.35 -> so agents can quickly and easily find the information
250.47 -> they need to resolve issues.
252.93 -> If we dive deep into the agent workspace,
256.14 -> we can look at a few of these capabilities.
259.02 -> On the left hand side of this screen,
261.72 -> you'll see the contact control panel where agents
264.57 -> can receive voice contacts, chat contacts and tasks.
270.51 -> An agent can also see if a user has been,
273.24 -> or customer has been, authenticated
275.07 -> using our voice biometric solution.
278.43 -> In the center of the workspace is our customer profile.
282.63 -> Here, customer information is aggregated
285.6 -> from disparate sources and displayed to the agent.
289.26 -> Information like prior contact history,
292.38 -> basic customer information, like name and phone number,
296.16 -> as well as other additional and custom information
299.07 -> the agent or the organization enters.
301.59 -> We also have another tab called Cases.
304.41 -> With Cases, organizations and customers can track,
308.22 -> collaborate on and resolve issues
310.71 -> that span multiple contacts.
313.47 -> In the Cases tab all the cases associated
316.11 -> with the customer profile are listed
318.24 -> and an agent can click into the case
320.76 -> or create a new case based on the issue.
324.45 -> In the top right of this workspace,
326.46 -> we have automatic recommendation service.
330.45 -> Amazon Connect can listen to the conversation
333.24 -> and automatically suggest knowledge articles
336.03 -> to help the agent resolve the issue faster,
338.97 -> or the agent can type a keyword into the manual search
343.29 -> to pull up a relevant knowledge article.
346.8 -> I just want to emphasize again that this agent workspace
349.53 -> is free out of the box and it organizes and unifies
354.21 -> all of our agent empowerment capabilities,
356.85 -> which you can use on a pay as you go basis.
361.2 -> - So now that we have an agent workspace,
363.36 -> we did ask ourselves what can we do more to help
365.97 -> and to identify three key areas
367.74 -> we wanted to help customers make improvements.
370.44 -> The first has to do with onboarding
372.3 -> and ramping up new agents and making them proficient faster.
376.23 -> The next has to do with making sure
377.88 -> that all agents, experienced or novice,
380.52 -> are adhering to best practices
382.71 -> and following updated standard operating procedures.
385.86 -> And lastly, we wanna help business operations teams
389.04 -> figure out new ways to improve efficiencies
392.04 -> and make sure their overall agents are operating well
394.367 -> and adjusting to new use cases.
396.87 -> And so with that, as of yesterday,
398.55 -> we've launched step by step guides as a preview feature
401.73 -> in our Amazon Connect workspace.
404.01 -> And so with this feature you're now able
406.41 -> to configure the primary tab of the agent workspace,
410.22 -> so at the start of the call,
411.39 -> you can surface relevant info and actions
413.82 -> to your agents to resolve issues faster.
416.91 -> At the highest of levels what we have done is
419.76 -> we've given you the ability
420.75 -> to surface contextually relevant data and actions,
424.14 -> making sure the agent has all the information they need
426.21 -> to get going quickly and the tools to dive in fast.
431.58 -> In particular when it comes
432.63 -> a little bit down deeper to a step-by-step guide,
435.42 -> here is where you're able to really structure your SOP
438.27 -> and codify into the system so an agent
440.28 -> is given clear instructions of what they should be doing
442.8 -> at any point in time when trying to handle
444.9 -> a particular use case.
447.21 -> So in terms of use cases,
449.19 -> it's gonna vary greatly by customer and line of business,
452.85 -> but at the highest level,
453.78 -> we do see four common steps that all contacts have,
457.41 -> and we really try to optimize the system
459.18 -> to make these steps work well.
461.49 -> The first has to do with identifying the customer
463.74 -> and making sure they are who they are.
465.39 -> And so with that, we have customer profiles
467.4 -> and voice ID plugins that make it a little easier
469.92 -> to get this step done quickly so the agent can dive fast
473.58 -> into the actual issue.
475.62 -> But first they gotta actually determine what that issue is.
478.29 -> And so with that in our landing page experience
480.27 -> we give you the tools you need to help guide the agent
482.85 -> to figure out what's going on in this contact
485.37 -> so that they can find the right action
487.14 -> to carry forth and solve it.
489.18 -> Now when it comes to resolving the issues,
491.34 -> which is probably the heart
492.27 -> of our step by step guides feature,
494.19 -> there will be various different approaches you can take.
496.71 -> And later on in the deck we'll dive deeper
498.78 -> into how this could work.
500.67 -> And lastly, most contacts require
502.77 -> some type of after contact work,
504.87 -> whether it's call notes or distribution codes.
507.27 -> And so with the guided experience we provided,
509.4 -> you'll be able to configure this in our agent workspace.
515.047 -> - All right, so let's see, step by step guides in action.
519.81 -> The video I'm about to show you,
522.09 -> there will be an agent working at an airline
525 -> who uses guides to help a customer
528.87 -> with a lost baggage reimbursement claim.
539.04 -> - [Voiceover] In this video, we will show
540.57 -> how the agent will submit a claim for lost luggage
543.06 -> on behalf of a customer.
545.16 -> When a contact comes in, we will display the match cards,
548.4 -> which is a quick snapshot of the step-by-step guides
551.1 -> that were configured by managers.
553.41 -> When agents click on a match card,
555.3 -> we will display the solve cards
556.92 -> with actions that will initiate a workflow.
559.68 -> The customer information is already pre-populated
562.11 -> from the Amazon Connect customer profile.
564.66 -> The most recent flights are available, and once selected,
567.63 -> the rest of the information is populated.
570.78 -> The agent asks the customer what items were lost
573.42 -> and the amounts that the customer wishes to get reimbursed.
576.66 -> The receipts can then be sent via email if needed.
580.68 -> In this step, the customer agrees
582.66 -> with the reimbursement amount.
585.34 -> Finally, the agent can review all steps with customers
589.08 -> to make sure all of the information is accurate.
592.17 -> And the form is successfully submitted.
600.418 -> - [Sandeep] All right, well let's call out a few things.
602.73 -> The screenshot you see here is step one of the guide
605.97 -> from the video we just watched.
608.61 -> Now, guides open up in the same tab as the landing page,
613.11 -> so an agent doesn't have to leave a contact.
617.01 -> Further, there is a progress bar
619.05 -> on the left hand side of the guide
621.068 -> and basic contact details on the top of the guide,
624.54 -> so the agent never loses context on the contact.
628.8 -> Information from first party sources,
630.84 -> like customer profiles, can be used to pre-populate forms.
636.18 -> Information from third party sources,
638.1 -> in this case, flight management systems,
640.74 -> can also be brought into the guide.
644.7 -> There are also navigation buttons at the bottom,
647.58 -> so the agent can progress through the guide
650.34 -> or cancel the guide altogether.
654.3 -> Now you've seen an example of how guides work,
659.82 -> and you're probably asking,
662.31 -> why have we designed guides in this specific way?
666.668 -> - And so we wanna take a small segue
668.49 -> to explain how we've used guides
670.23 -> in our own Amazon customer service departments
673.38 -> in order to improve agent efficiency and training.
677.856 -> And so our old custom CCP looked like this,
681.69 -> and as you can maybe tell,
682.92 -> there's a lot of features for agents,
684.78 -> which is great because we gave them
686.04 -> all the power they needed to resolve any type of contact.
689.31 -> But as time went on and we added more features,
691.74 -> and use cases got more complex,
693.57 -> this started to become a burden on agents
695.28 -> because it became a very unwieldy system to use.
698.25 -> So for example, just to look up
700.14 -> the status of a Prime subscription membership,
703.11 -> agents would have to go through three menus
705.54 -> and four UI pages just to get the status details.
708.45 -> And then to actually modify the subscription,
710.61 -> they'd have to do that three more times
712.59 -> going through 15 different UI pages and many menu tabs.
716.16 -> And overall this is becoming too burdensome of an experience
719.49 -> that is placing high cognitive load on our agents
722.73 -> that we've noticed was reducing their ability to learn.
725.73 -> And this is true for new agents just trying
727.56 -> to learn the system and get acquainted with their job.
730.41 -> But even experienced agents we realized
732.33 -> were diverging in how they handled
734.49 -> the same type of customer requests
736.2 -> because the system was so fluid and open,
738.6 -> and that was leading to inconsistent customer experiences.
742.26 -> So we dug deep and we tried to evaluate
744.384 -> what is really going on with our UX.
747.09 -> And when we evaluated it, we realized our UX process
750.27 -> was asking the agent to listen to the customer,
753.33 -> simultaneously locating and researching data
757.14 -> and buttons and menus,
758.88 -> and then on the fly assembling a solution to that problem.
763.53 -> And this was wasting a lot of time
765.24 -> and a lot of energy for our agents.
767.55 -> So after extensive UX research and many, many prototypes,
771.87 -> we've landed on a new UX process
774.666 -> and we call it listen match solve.
778.14 -> The idea is that the agent will primarily focus
780.48 -> on just listening in the customer,
782.43 -> and we're gonna surface on the UI screen
784.359 -> the most likely topics for that call
786.66 -> so that the agent can just hear what the customer is saying,
789.78 -> click on what they see on the screen that matches the topic,
792.39 -> and then proceed to dive right into a solution.
795.15 -> So we're gonna illustrate this by walking through our UX
797.46 -> with a little bit more detail.
799.8 -> So in this UX, we surface the most likely topics for a call.
803.7 -> And a topic could be a product order,
806.82 -> it could be even a financial product,
808.56 -> it could even be a workflow.
810.66 -> Either way, we give many options to the agent
813.36 -> so that one of them is likely the issue.
815.61 -> And once they click on the topic,
817.17 -> they can see a summary of pertinent details about that topic
821.28 -> as well as actions that let them update this topic.
825.18 -> And what we found is it's only so many actions
827.969 -> you can do for a particular topic,
830.97 -> only so many things you could do for a retail order.
833.52 -> And so by narrowing down the agent position tree
835.56 -> to first just find that right topic,
837.81 -> it really funnels down the options they have to choose
840.42 -> between really strictly limiting their cognitive load.
844.38 -> But now they found the right topic, the right issue,
846.99 -> how do they solve it?
848.34 -> And that's where the step by step guidance really kicks in,
850.86 -> where we give these sequences of simple UI pages
854.13 -> to the agent to really let them focus
855.96 -> on one thing at a time, whether there's one input field,
859.62 -> one question to ask the customer,
861.36 -> they can slowly go step by step,
862.92 -> getting the right information they need
864.78 -> to help the customers matter.
867.24 -> Additionally, along the way,
869.1 -> we will feed a bit of agent scripting to the agent
872.07 -> so that, let's say they submit a transaction
874.32 -> and they get a confirmation message,
876.03 -> it won't just say that it's been successful.
878.04 -> The success message will read as a script
880.23 -> they can read out loud telling the end customer
882.63 -> what's just happened and what to expect,
884.64 -> like an email notification.
886.5 -> Additionally, even when doing different questions
889.47 -> and inputting fields,
890.73 -> with inline standard operating procedures and policies
894.21 -> right into the question, so in case a a customer
897.3 -> has a question on the spot about different options,
900.15 -> the agent can read them right out loud,
902.31 -> rather having to go find a knowledge article
904.77 -> in some buried data.
906.96 -> Now let's say our topics were wrong.
909.18 -> Well, in that case,
910.013 -> we always give an escape hatch to the agent
912.72 -> to find the right workflow for them.
914.43 -> So for example,
915.263 -> if the agent clicks on the find more button here,
917.46 -> they'll see a whole list of workflows
919.65 -> that they have access to.
921.06 -> And then now they can figure out,
922.087 -> "Okay, which one do I need?"
923.58 -> This way we give them the freedom to always self-serve
926.19 -> and find the right solution in case our prediction
928.23 -> is a little bit off.
930.24 -> So now I wanna talk about the outcomes
932.16 -> of this new UX we've rolled out.
934.35 -> First, we've been able to reduce onboarding time
937.71 -> by down to 50%.
939.99 -> And the main reason for this is the app is so intuitive
942.72 -> we don't have to spend that much extra time
944.49 -> teaching all the different menu layers.
946.62 -> It's pretty much a two different panel screen to go through.
950.46 -> But more important, we've actually been able
952.77 -> to reduce our time to proficiency by 40%.
956.67 -> Agents are using the same UX
958.29 -> regardless of the use case they follow,
960.36 -> so it really creates a simple streamlined experience
963.21 -> so they can learn as they go.
965.13 -> It's the same methodology of listen, match, solve
968.13 -> regardless of the new line of business
969.81 -> they might be assigned to, or any use case.
972.27 -> So they can basically learn as they go
974.4 -> and not have to worry about doing retraining over and over.
978.42 -> Additionally, we've been able to reduce
980.13 -> our average contact handle time by up to 35%
983.58 -> for some departments.
984.75 -> And right now we're at a global average of 20% reduction.
989.22 -> And perhaps most importantly,
990.54 -> we've actually reduced agent heart rates.
992.82 -> And I take this really seriously.
994.05 -> We did do biometric studies
995.7 -> where we hooked our agents up to devices and did AB testing
998.91 -> of the old UX comparing them to new UX,
1001.85 -> and the heart rates did go down
1003.8 -> and self-reported stress levels also reduced.
1006.59 -> And while we don't advocate Connect
1008.45 -> as a healthcare product in this manner,
1010.94 -> we are really just happy that we're able
1013.19 -> to make the daily lives of our agents better.
1018.83 -> - All right, well this UX process was research and developed
1022.94 -> in the context of Amazon customer service.
1026.51 -> But we can really generalize this UX process
1029.15 -> and apply it to other types of organizations and use cases.
1034.22 -> Here we have more of a generic retail organization example.
1038.87 -> An agent at a retail organization is using guides
1042.92 -> to help a customer return address.
1046.64 -> The customer's information, purchase history,
1050.45 -> as well as payments, are all consolidated
1053.93 -> into a single workflow for the agents.
1057.53 -> Can we go beyond retail?
1059.9 -> What about financial services organizations?
1063.56 -> Well, in this example,
1064.67 -> an agent at a credit card company is using guides
1069.47 -> to help a customer add an authorized user to a credit card.
1074.12 -> New user information entry, approvals and follow ups,
1078.5 -> again are consolidated into a single workflow within guides.
1085.4 -> How about hospitality?
1087.65 -> In this example, an agent working for a booking site
1091.97 -> is using guides to help a customer with a new reservation,
1097.16 -> whether it's one reservation, two reservations,
1100.37 -> handling payments, it's all consolidated
1103.28 -> into a single workflow within the guide.
1108.08 -> Now, guides are not just for voice contacts,
1112.1 -> guides can be used for chat contacts
1115.97 -> and the appropriate guides are surfaced to the agent
1119.96 -> based on the context of the chat.
1124.4 -> But it's not just voice contacts or chat contacts,
1128.319 -> you can also use guides with Tasks.
1132.08 -> Tasks can be used by agents to create, assign
1137.24 -> and complete tasks in the same user interface
1141.08 -> that agents accept calls and chats.
1144.62 -> Again, the appropriate guides are surfaced
1147.38 -> based on the context of the task
1150.05 -> that is received by an agent in a queue.
1155.6 -> Guides can also be used by agents
1157.61 -> who are in after-contact work mode.
1160.88 -> You can see here the contact control panel on the left,
1164.75 -> the agent is in after-contact work mode
1167.24 -> and a new tab has popped open
1169.37 -> so the agent can complete a feedback form
1172.91 -> as part of their guided experience.
1177.26 -> Now, the default design that we've
1179.3 -> been showing you throughout this presentation
1182.36 -> was really inspired by Amazon customer service.
1185.87 -> And you see some of the successes and outcomes
1188.3 -> that Jack showed you previously.
1190.67 -> But we know that customers want customization.
1195.171 -> And with guides you can customize the way they look
1198.56 -> inside of our Amazon Connect agent workspace.
1202.362 -> In this example, you see that there is a guide
1205.57 -> in the central panel and we've also persisted
1208.73 -> all of the available guides in the right hand panel,
1211.85 -> in case an agent wants to make a switch.
1217.55 -> Going even further, guides are flexible.
1222.17 -> You may want to take guides out
1223.88 -> of our Amazon Connect agent workspace and embed them
1227.15 -> in your custom agent workspace, or in your CRM.
1232.94 -> We know that you make your build decisions
1235.31 -> and your purchase decisions to fit
1238.13 -> the unique needs of your agents.
1240.35 -> And you can use our APIs to bring a guided experience
1244.28 -> into a custom workspace like you see on the screen here.
1248.6 -> You can see here that we have
1250.49 -> the contact control panel on the bottom right,
1253.25 -> and we have other widgets,
1255.18 -> like customer information on the bottom left,
1257.87 -> queue statistics on the top left,
1259.85 -> and then customer sentiment scores on the top right.
1263.57 -> Now whether you use guides
1264.95 -> in our Amazon Connect agent workspace,
1267.35 -> or you use them in a custom agent workspace,
1270.35 -> the way you build them is exactly the same.
1275.21 -> - So I had just a question,
1276.26 -> how can customers build guides?
1278.36 -> And when we set out to build this experience,
1280.31 -> we could have released a bunch of APIs,
1282.8 -> but we really wanted to go with
1283.97 -> a more low code, no code approach
1286.22 -> that enabled any persona in a contact center organization
1289.91 -> to build these guides.
1291.26 -> And when thinking about how to do this we realized
1293.54 -> we already have a no-code, low-code builder in Connect,
1296.06 -> and that's Amazon Connect flows.
1298.1 -> With flows, which launched with Connect five years ago,
1300.62 -> you're able to use this drag and drop GUI to build IVRs,
1303.98 -> chat bots and routing logic for contacts.
1306.98 -> But now with the addition of
1308.18 -> one new flow block called the show view block,
1311.39 -> you can create UI-based workflows to surface to users.
1315.44 -> And so with this flow block what happens is
1317 -> we send instructions to a client application,
1319.76 -> in our case the agent workspace, determine what to render,
1322.7 -> and then based on the agent's actions and what they click,
1325.04 -> a response is sent back to the flow
1326.87 -> and then it carries on executing the flow,
1328.88 -> either doing backend updates
1330.92 -> or going to the next view to show.
1334.28 -> This flow block can be combined
1336.05 -> with any other existing flow block
1338.24 -> in order to create branching logic
1340.43 -> as well as get customer data.
1342.89 -> So in terms of integrating data,
1344.588 -> both options are available to both integrate with Connect
1348.02 -> as well as get data from outside of Connect.
1350 -> So in terms of data that already lives in Connect,
1352.07 -> it's all that IVR data just ready to be used.
1354.47 -> This can be surfaced at screen pops
1355.94 -> true agent at the start of the call,
1357.53 -> but also used as conditional logic
1359.84 -> to figure out where to branch in the workflows.
1362.63 -> Additionally, you can use our
1364.28 -> out of the box cases flow block
1365.92 -> to get case information to an agent
1368.24 -> and even automate the creation of cases.
1370.82 -> An agent can go through a step by step guide and if
1373.01 -> your business logic determines follow up work is needed,
1375.77 -> a case can be automatically created on their behalf,
1378.2 -> tasking around to other agents without the agent
1380.257 -> having to do the manual or tedious work
1382.58 -> of all that data entry.
1384.5 -> And additionally, likewise,
1385.82 -> customer profile data readily available to the agent,
1388.82 -> both you and their guides,
1390.86 -> as well as to give them the ability
1392.48 -> to create new customer profiles for new customers.
1398.57 -> In terms of integrating data from outside of Connect,
1401.33 -> we use the Lambda flow block
1402.89 -> that already exists in flows today.
1405.62 -> Traditionally used to do data dips in IVRs
1408.428 -> as well as in automated tasks,
1410.66 -> the flow block can now be used
1411.83 -> to get data from external systems,
1413.93 -> whether it's a vendor product or homegrown solution,
1416.93 -> and then pull that into the agent UI
1419.15 -> to give them information they need.
1420.71 -> But also a way to update these backend systems
1423.44 -> at the end of a transaction.
1425.628 -> And what this looks like in practice is
1427.97 -> all this data from disparate sources
1429.95 -> is unified into single UI pages.
1432.65 -> This way the agent isn't worried about
1434.39 -> where customer profiles data lives,
1436.19 -> or different transaction history.
1438.56 -> They just get one tool to make the updates.
1441.32 -> And to them you remove this burden
1443.09 -> of figuring out where they need to go.
1446.994 -> So there's a few other benefits of using flows
1448.7 -> for the solution that we wanna call out.
1451.37 -> The first is ability to transfer contacts between agents.
1455.39 -> If Agent A is working through a workflow
1457.76 -> and they get stuck or they need authorization
1460.01 -> from their manager to complete it,
1461.66 -> they can transfer contact to agent B,
1464.12 -> and agent B in their screen will be able to see
1466.682 -> how far agent A got and then carry on with the work.
1471.44 -> Additionally, you can create branching paths
1473.51 -> within your guides that way there's complex business logic
1476.729 -> or standard operating procedures.
1478.61 -> It can all be codified right there in the system
1481.28 -> that way the agent doesn't have to worry about
1483.14 -> reading up knowledge articles and rules.
1485.27 -> It's right there for them
1486.41 -> and guiding them throughout the process.
1489.47 -> Additionally, there's a linkage
1491.36 -> between the workflows agents do through all flows
1494.33 -> and the original inbound contacts.
1496.28 -> This allows us to update the custom attributes
1498.62 -> associated with an inbound contact,
1500.053 -> such as a distribution code,
1501.89 -> giving you the way to really sync your data.
1504.571 -> And on that vein,
1506.21 -> we also will have audit trails of agent actions.
1509.3 -> If you go now to a voice contact or chat contact record,
1512.93 -> you'll be able to see all the different workflows
1515.12 -> that an agent performed during that contact.
1517.85 -> And when you click on a workflow,
1519.29 -> you'll get the complete audit history
1520.64 -> of every click they took, the steps they performed,
1523.25 -> and even the data they're trying to update.
1527.24 -> So I do wanna reiterate that a guide is really just a flow.
1531.68 -> It's a knowable contact flow,
1533.3 -> but it just has this capability of sending data to a UI
1536.54 -> and determining how that UI should look.
1538.73 -> But there's always benefits of sharing the data
1540.62 -> between the different types of flows
1542.39 -> for transferring between different contact flows,
1547.19 -> which will help for both the maintenance
1550.01 -> and modularization of your workflows.
1552.137 -> And additionally, you get the benefit of being able
1554.12 -> to use flow modules to really create discreet units of work
1557.96 -> and logic that keeps the flow maintenance scalable.
1562.1 -> But in very particular note
1563.39 -> on the serve connection between flows,
1565.46 -> the way you actually determine
1566.66 -> what that first screen should be for an agent to see
1569.54 -> has to do with setting the immediate default flow
1573.23 -> at the server contact.
1574.73 -> So when you go to your contact flows for routing,
1577.67 -> you'll be able to set one contact attribute
1579.92 -> that determines which guide to surface.
1581.81 -> And you can do this based on different branching rules.
1584.54 -> You could decide, okay, for certain queues,
1586.58 -> this is the right landing page to show.
1588.5 -> Or maybe depending on the IVR response,
1590.72 -> you dive right into a problem.
1592.64 -> So for example, you'll be able to either in the situations
1595.46 -> where you have no idea what the call's about,
1598.31 -> you can adopt our listen match solve UX
1600.471 -> and give the agent a bunch of topics,
1602.3 -> then figure out, okay, what should I do next?
1604.7 -> But in the situations where you already know
1606.38 -> the exact intent from the IVR, customer history,
1610.1 -> you'll be able to surface the exact flow
1612.35 -> and immediately the step the agent needs to do
1614.78 -> right on their screen.
1616.25 -> This way you get full configurability
1618.95 -> into how dynamic you want to make this experience
1621.05 -> and to really drive down the agent contact handle time.
1625.64 -> The last slide we have
1626.69 -> about flows and the configuration experience
1628.67 -> has to do with analytics and flow logs.
1631.85 -> Basically, we've updated our flow logs
1633.74 -> to now support the tracking of agent actions.
1636.77 -> So you're able to first look at a particular contact
1639.56 -> and see as to which guide they chose,
1641.96 -> and then which steps along the way
1643.7 -> and given different decision trees do they take.
1646.34 -> But then more so you can view all the different workflows
1649.58 -> executed in aggregate as figuring out, okay,
1651.89 -> what percentage of my contacts
1653.42 -> did the agent take this particular workflow,
1655.55 -> and then down the way, how often did this branch get chosen?
1658.87 -> Overall, by given this data,
1661.4 -> we really hope to allow managers, business analysts,
1664.28 -> and whoever is creating these workflows
1665.74 -> to get the insights they need to optimize and improve
1669.17 -> the workflows by spotting trends as well as sticking points.
1672.62 -> Maybe there's a particular step
1673.971 -> that there's a system error and it always errors out.
1677.99 -> Or perhaps agents are manually abandoning
1680.09 -> the flow of that step indicating a UI problem.
1683.12 -> Either way, we want to get the data
1684.62 -> right to the manager's hands with some
1686.24 -> helpful pre-canned cloud watch queries
1688.58 -> that really allow these personas
1690.92 -> to update these flows efficiently
1692.9 -> and make sure they're more effective for agents.
1698.78 -> - Okay, well, customers who use Connect today
1703.34 -> are excited about step by step guides.
1706.91 -> One such customer here, Unum,
1709.461 -> helps millions of people every single year
1713.45 -> gain affordable access to life, disability, accident,
1719.471 -> illness, dental, and vision benefits via their workplace.
1726.23 -> Today, they use a custom agent workspace
1730.25 -> that their agents have to help customers.
1733.34 -> They're excited about guides to consolidate actions
1737.48 -> into a single workflow and take the number of screens
1740.78 -> their agents use from five down to one.
1745.52 -> Further, they believe they may be able
1748.37 -> to sunset their existing workspace
1751.28 -> and save 15-20% on maintenance costs
1754.49 -> with the Amazon Connect agent workspace.
1758.9 -> Partners are also excited about guides.
1762.62 -> Voice Foundry specializes in
1764.78 -> the design and delivery of customer experience solutions
1769.37 -> built on the Amazon Tech Stack.
1772.94 -> They're excited about the flexibility of guides.
1776.78 -> They believe it'll help them deliver agent experiences
1780.35 -> more rapidly in an evolving marketplace.
1785.93 -> So that ends the main part of our presentation today
1788.75 -> on step by step guides.
1790.64 -> If what we've been talking about today
1792.378 -> has interested you or excited you,
1795.2 -> Jack, myself and our colleagues
1797.42 -> will take questions around the floor
1800.96 -> after we end the presentation.
1803.54 -> If you are new to Connect
1805.1 -> or are interested in learning more about Connect,
1807.41 -> here is a list of other sessions we have today and tomorrow.
1811.97 -> I'll leave it up there for a second
1813.11 -> if you want to take a note on a particular session.
1820.4 -> Otherwise, thank you.
1822.08 -> Thank you so much for joining us today.
1824.261 -> If you wanna reach out to Jack or myself,
1827.739 -> we have our email addresses up here,
1829.91 -> as well as a QR code you can get our contact information.
1834.35 -> Otherwise, have a great rest of your day.

Source: https://www.youtube.com/watch?v=wDJ0h8vu-RQ