Roundtable Review: AI and Automation in the Workplace

February, 22 2021

Subscribe and stay up to date

No spam, we promise! You will only 
receive essential emails.

Every day we are seeing more and more AI and Automation being implemented into businesses to solve issues not only internally but externally as well. To get the most out of AI, businesses must understand which technologies perform what types of tasks, where they need to prioritize AI in their business, and develop a plan to implement. Automating business processes, gaining insight through data analysis, and engaging with customers and employees are all ways in which AI can help improve your business, but where do you need it the most? How can implementing AI increase productivity, efficiency even culture within your business? We sat down with leaders in the CX industry to not only answer these important questions but to also give insight into their own journeys implementing AI into their businesses. 



Simon Choi- Technical program manager at Paypal, working for the localization group supporting more than 50 languages but in the process of getting to 100. In charge of processes and tools, he manages important tools that the team relies on, such as Smartling, Talla, and Jira. 

Deb Clow- Director of project management for MyPoint Credit Union in San Diego. Duties include keeping projects on track, finding new ways of streamlining efficiencies, and looking for ways to make information more attainable to not only employees but down the road to members as well.

Greg Peppel-  Works for Avero as Head of Products and Customer Support. Avero is a reporting and analytics application for the restaurant industry, building web and mobile applications that help restaurant managers understand how their business can perform better.



Can you tell us a little bit about how you're trying to use AI today to improve your customer experience and what problems or improvements are you hoping to solve by using AI?



Simon- “At any given time we have close to 300 translators working for us and it is very fast-moving. We send out more than 20,000 or 30,000 new content for translation that should be translated into more than 50 languages and then we get them back within a week so it’s really fast-moving. In the past it was very hard to provide correct information at the right time, so we have a lot of documentation on product terminology like PayPal products, there are so many PayPal products as well as processes. So all these people working on translations they need almost immediate answers otherwise they cannot move on to the next stage. Now using Talla it's very easy to find the knowledge that they need to move forward with their translation. What is also really good about this is in the past we would get the same questions from like 50 different translators, but now because of Talla if you don't have the answer, then you can instantly create a ticket then provide that information to one person and that can be shared by everyone else so we don't have to deal with replying and asking repetitive questions.”


Greg- “We are putting AI at the front lines of that support experience so we are using the customer assist chatbot from Talla to basically field inbound support requests from customers and ideally that's the first interaction that a customer has. The goal or why we chose to set out to make that change and put them at the front lines, again looking at the impact of the pandemic on our business we just weren't able to keep as large of a support team as we wanted, we didn’t have as many customers for a large part of 2020 and so we needed to figure out a way to suppress the number of tickets getting put in front of our agents and it has definitely been able to do that substantially. The unexpected consequence of that is actually led to a better customer experience because the average time that it takes a customer to get a good answer from that chatbot is seconds vs when it ends up in our ticket queue there's going to be one to two-day turn around before we ultimately get an answer for that customer.”


Debbie- “We had recognized quite some time ago we had a problem with procedures, we had them everywhere. We had a local drive, everyone had them on their own computers, so we really realized that we needed one truth, one source, and a place where our front-line staff could get an answer to a question or a procedure quickly and efficiently which is how we came to implement Talla. We are using the chatbot for just our internal teams right now but we've really been able to identify and with confidence say now we have one truth and one location for procedures it has definitely streamlined the ability also to put some dates and names to who's going to review the procedures and with what regularity were going to have that done. We were a little bit of a wing and a prayer before as to whether or not that was happening so our audit department is definitely impressed with the structure that we have set up on it now but AI has really taken what used to be quite a challenge to find an answer especially for new employees and made it a very simple process.”


What suggestions would you give to companies who might be exploring implementing AI and putting it into their CX?



Simon- “When we decided we needed a knowledge management tool we actually reviewed many other tools too but when we did our ROI analysis we did a survey and then looked into how much time translators and internally we spend looking for correct information. Internally we never know if this information is out of date, if it’s correct, or even if there are multiple versions of the same information. We waste a lot of time looking for correct answers, so we did all this ROI analysis, and when we did the evaluation of these tools Talla came out at the top because it has the best AI. Other tools may be more mature like they may have already developed plug-ins and things like that but when we did the actual AI test Talla came out first. So I really strongly suggest looking at the time lost and you may be surprised because came up with some phenomenal dollar amount, and purchasing Tallas services were a fraction of the money.”  


Deb- “I would highly encourage the test teams, we had a test team that did a lot of work for us in the very beginning asking questions and really helping us identify strengths, weaknesses and get a true comfort level and true subject matter experts out there that could help answer follow-up questions once we did go live with everyone.”


Greg- “So we did a phased deployment, so we didn't release this chatbot to all of our users simultaneously. We were able to release it in waves almost so we started with very few numbers like 10 customers. We didn't really have a good idea of exactly how people were going to phrase questions or exactly what those questions might be and we also felt like you only have so many opportunities. If a customer goes in and they have a bad experience with that chatbot we thought it was unlikely that they would come back and expect that experience to be drastically different. So it was very important to us to control the rollout and not ultimately deploy in a widespread way until we felt like it was fine-tuned enough to actually create a good experience out of the gate for a customer.” 


What excites you about AI in the future, where do you see the future going with AI and customer experience in particular? 



Deb- “I have been in the financial industry for a very long time so this pandemic has definitely made us stop and rethink how do we do business. Members are not coming into the branch as often so how can we still get our message out to the members? So we are looking at doing some marketing automation this year which I'm kind of excited about and just really figuring out how to still be successful in the space we are in without the face to face contact with our members but still having that connection to them even though we don't see them every day like we used to.”


Greg- “I am really interested in thinking about how AI can create structured data out of unstructured data. When we give users information about their performance we're getting that underlying data from 60 different POS systems across our customer base and the data structure for each of these systems is different so there's a very manual process that can go into aligning that or it just remains unaligned, but you can start to think about how an AI can identify the data and dimensions and put it into a uniform structure. It makes it easier to build a broad-based application, then that can sit on top of that clean data layer and be more insightful for those customers so it's not just using the AI in the Direct customer experience but using it to structure the data that supports a good customer experience.”


Simon- ”We are looking into the possibilities of integrating Talla with our other internal tools. So we have our internal PayPal instance with Jira so if we could connect Talla together we would be able to turn any request or piece of knowledge into Jira tickets for the resolutions or to share with other members of PayPal not just with localization but outside our team, then we can collaborate with other people then turn the knowledge back into our Talla. So that automation is something we are looking forward to having this year.”


Technology is working as intended but has the rollout improved the tone or culture within your organization? Has it helped employee engagement? 



Deb- “I can take it from the standpoint of it has definitely helped improve our consistency in training. We have used Talla to identify some core procedures based on what position the person is that everyone needs to review in 30- 60- 90 days, so we have a comfort level now that the consistency of that process is happening. We’re hearing from our frontline team that they feel more empowered because they can get the answer with a member either on the phone or in front of them if they happen to come in quickly and they don't feel like they’re waiting around for someone to get back to them or having to get back to the member. So we have seen a positive improvement within kind of our tone with our employees. Since we rolled this out the ease of doing business for them a whole lot easier and somewhat improved the communication between training and I think sometimes they were afraid to ask a question because they didn't want it to come across as maybe they should have paid attention to something that they didn’t so they can ask the chatbot and feel a little protected in whatever their question might be. So we have seen definitely a positive increase from that.”


Greg- “It shifted our focus from a support perspective in the sense that before we implemented AI on the front lines of our support the mission of that team was to triage customer issues as quickly and courteously as possible. Now I would say the mission is to continue to invest in the knowledge base to prevent issues or tickets from ever being created and yes we still have to do triage and we still have to manage tickets but that's not the primary sole focus of the support organization anymore and I think that's been a really positive cultural shift. So the time you save in managing the support queue we get back as an investment in the knowledge base which is ultimately better for the customer. Reps are constantly saying I don't want to answer the same question over and over and over again isn’t there a bot that can do this? That’s an expectation people have now, they want to be put to a higher purpose and I think that's what investing in our knowledge base is for them. We almost think of our knowledge base as a feature of our product and we are measuring and putting investment into it and the support agents are almost our engineers for that knowledge base now.”


I’d love to learn about how you chose Talla among other brands in the market? Do you have any tips for people as they start evaluating solutions?



Greg-  “Talla understands that vicious cycle of repetitive Q&A and I think puts that front and center into how their product is designed. It recognizes the importance of the relationship between the knowledge base and the chatbot and it almost forces you to reinvest your energy into improving the knowledge base and I really like that. There were other impressive chatbots but I don't think they would have had this forcing mechanism of getting you to constantly reinvest in your knowledge base.” 


Deb- “I would encourage you to ask whoever your rep is or whoever is doing the demonstration for you for some references. We did spend a lot of time talking to references and asking those day-to-day questions, how are you using this, what have you seen as an improvement, and what advice would you give someone that's looking to bring this into their environment? I can tell you that the references from Talla really helped us make the decision to go with this solution, they helped reinforce that the real everyday benefits can be seen.”


We hope you enjoyed our question and answer portion of our Roundtable Virtual Discussion! Thank you to everyone involved, we hope to have more talks like this in the future so we can answer even more questions you might have on AI and Automation!

View all posts

Subscribe and stay up to date

No spam, we promise! You will only 
receive essential emails.

Subscribe and stay up to date