Webinar on Business Continuity Planning With AI in times of Covid-19 & WFH


Business Continuity is the biggest challenge for all the companies in the world, whether it is small, medium or large scale industry. The same is true for local or multinational firms. These are the times which businesses have faced in their lifetime. This is not a local phenomenon for which companies could have their BCP in place.

Things need to be done immediately without losing any time. Do we have any options? Can automation help businesses in such the Covid-19 era? Can Conversational AI and Chatbots help businesses?

Looking for an answer? Raghu Ravinutala, CEO & Co-Founder, Yellow Messenger, shares his experience and solutions which Yellow Messenger conversational AI platform can provide to businesses in matter of a week or two.

Below is the Transcript of the webinar

Host: Good Afternoon Everyone. Welcome to the webinar titled “Business Continuity Planning with AI in times of Covid 19 and WFH (Work From Home)”.

On behalf of NASSCOM, first of all I’d like to welcome you all and thank you for taking your time to attend this webinar. To introduce myself, I am Hitesh Achra, part of the Northern India membership team at NASSCOM. 

Our key presenter today is Raghu Ravinutala, CEO and Co-Founder of Yellow Messenger. In today’s webinar, it will cover how every company big or small who are grappling with an influx of queries from customers and employees given the Covid-19 pandemic can learn executable strategies to turn around the BCP threat using conversational AI. We will take questions and answers at the end of this session. you may put in questions in the questions panel of the GoToWebinar console. I now request Mr. Raghu Ravinutala to take over from here. 

Raghu Ravinutala: Hello everyone, Good Afternoon. Hopefully, all of you are having a productive time working from home and getting to spend more time with families.

l I’ll give a brief introduction about Yellow Messenger, so that it sets the context through this entire webinar. Yellow Messenger is an enterprise conversational AI platform. It’s been four years since we have been working with 100-plus enterprise customers globally. We have

been working with enterprises to automate everything from customer engagement towards employee engagement, ITSM, HR automation, etc and I think we are we’re proud that we’ve been able to create 100 million dollars of impact through these automation deployments across these enterprises. 

The reason we are doing this seminar/webinar is over the last three months we’ve been engaging with multiple enterprises and especially over the time where we have seen Covid-19 started impacting, a lot of enterprises have come back to us and asked about how conversational AI can help them solve these problems. 

We have seen four or five enterprises successfully deploy conversational AI and we feel that it is useful for all of us to know more about how conversationally AI can impact the business continuity process across a wide set of enterprises. 

Maintain Business as Usual in Uncertain Times

So life has completely changed over the period of March 2020 and all of us are looking to maintain business as usual and calling these uncertain times, is actually not representing the real picture. It’s actually disrupted times. Things have come to fold where we haven’t imagined where we will not be able to access our office across our premises in the world.

So there is no redundancy across geographies, there is no redundancy across IT services, working across multiple places. So these are absolutely uncertain disrupted times and each one of us in enterprises and businesses are looking how to manage our businesses and provide products and services to customers at an acceptable level. 

Lockdown Across India and World

Indians Under Lockdown

There is a lockdown not only across India but across the world. A lot of GCC’s would be working across multiple geographies and we are seeing our customers as well trying to grapple with the situation where IT support helpdesk out of India, is not able to cover for disruption happening in the UK or France and these are the situations that BCP was not really planned for. A lot of companies and customers are opting for technology to solve their problems.

It’s kind of reflected in all the experiences that we personally are going through be it Amazon to Big Basket to insurance companies, where they are seeing significant volume increases in customer inquiries either due to increased demand for their services or increased number of queries due to uncertainty about some of the services whether my bank account will be available, whether my ATM would be operational, whether my VPN would work, in this scenario whether I need to move the laptop back to my home.

So are tons of queries and lot of time the constant response is that, “hey we are running on limited staff and will not be able to solve this problem currently. This is definitely taking the customer service multiple notches down from what businesses were providing in their regular scenario. I came across this really nice video and the reason I wanted to put it out is that we always relate BCP to something that’s there in the contract but don’t really understand why BCP is really needed and what is the definition of business continuity process.

So I’ll just play this video it’s about three minutes and I will come back to you after the video finishes. By the time, I think all of you should have a clear understanding of what is BCP and what it takes to provide that and why it’s important for the business.

Video Transcript:

Business Continuity Planning and Analysis

Business Continuity Planning Analysis

Ah! Wasn’t that a very simple representation of what we need to do as businesses. I would like to reiterate that point of the ability of the business to provide the products and services at acceptable levels and of course, we are in a destructive cycle and that’s the key focus of business continuity plan. If you see a business continuity planning, it’s addressed in three ways.


  1. Environmental
  2. Accidental
  3. Deliberate
  4. Secondary

There are threats that you typically see a company or a business faces and these threats could be different for different sets of organizations. Whether organizations are offline centric or the organization’s are purely digital businesses and these threats clearly require different strategies to address physical threats, they need different strategies. Digital threats need different strategies and all of these have vulnerabilities or single points of failures which need to be addressed.


  • People
  • Systems
  • Infrastructure
  • Equipment
  • Supplies and Stock
  • Knowledge

These failures can happen across people’s systems, infrastructure, equipment, supply chain disruption but at the core of it, if you see a lot of disruption where it happens is allowed how people have information to react around these scenarios. 


  • Cloud Transformation
  • Democratization of information
  • Automation of Processes
  • Unified Communications

A lot of business continuity planning in a regular course would involve redundancy of people, would involve redundancy of systems, would involve redundancy of supply chains in different environments. But it always needs people to interact and enable these different fallback mechanisms while addressing.

How do we Ensure Covid-19 doesn’t interrupt business

How do we ensure covid-19 doesn't interrupt business

Today,  we are actually faced with a unique situation with Covid-19. A lot of redundancy mechanisms that we have plans for in our regular basic course are no longer sufficient to address the continuity of our business. There are redundant workplaces but it’s locked down across the world, so people really are not able to go back to their offices. People are working from home. 

There are redundant call centers. So there is a call center in X location, there is a call center in Y location as a written it process but unfortunately we are not able to invoke that redundancy as it’s a long term across.

There are health issues where there is a real risk of one of our employees or colleagues being affected by a health reason due to Covid-19. That’s not a typical BCP risk that’s typically covered across the organization. It’s probably covered at the executive level but not at a really organization-wide level. 

So all of this is kind of putting people at the center of the process. How are we able to augment the limited resources that we have? There are people who are working from home or they have limited access to the systems. They have limited access to their colleagues, collaboration at the right time.  How are we to enable that limited set of resources to either address the increased inquiries that are coming in due to increased demand or increase our uncertainty around the products and services for your specific company.

 What if this was to happen?

I’ve seen a lot of articles around a lot of digital transformation being driven due to Covid-19 but why is it really happening? 

What if IT support automation at the company, whenever you’re seeing this increased volume of inquiries,  around how work from home and how my VPN works. If your ITSM and HR systems can start automating 50 to 75 percent of those queries automatically on your team, on your slack, on your telephony, the existing staff or existing experts are 10x more productive during these times. 

What if your call center or what if your customer support agents are augmented with conversational bots that can answer 60% to 80% of customer inquiries, completely reduces the burden on the existing workforce. That is providing customer service and this is actually possible and we are seeing significant examples of it. 

One of our largest Non-banking financial services customers, one of our largest insurance customers, has completely closed down their call center operations because people are working from home or just they can route calls to the agents and replaced it with customer support on WhatsApp business account where consumers can log on to the official channel, ask various questions about insurance, get health insurance, inquire whether Covid-19 covers their insurance policies, etc. And this has been so seamless that post the Covid-19 situation, the company is in fact planning to never reopen the regular call center again.

So that’s digital transformation in 1-2 weeks, that typically takes about a few years to achieve digitization of call centers. Conversational AI, we believe, is uniquely positioned to address the challenges that we are facing. The challenges that we are facing is essentially orchestration of processes across the organization be it customer support or be it IT support or HR support involves orchestration across people. People need to find out from multiple

people about what’s happening on a certain specific item, it means is orchestration across processes. So the data need to be fetched from multiple or different applications trigger workflows to provide service back to the customer and seamless communication to both the employees and customers about what are the changes that are happening within the company. 

We are seeing companies where the HR policies are getting changed. There are sick leaves, their paid leaves being introduced, so that’s the communication that needs to be addressed to the employees.  we are seeing companies where insurance companies need are creating new policies and helping their customers who hold existing policies in a way that can help them and communication is the top priority for them on channels that those consumers are or being available at this point in time.

One technology that can address across these spectrums of processes right from being able to interact the fronted and being able to orchestrate processes at the backend is uniquely positioned to solve this problem and we are seeing that happen in person.

60% of the conversations that enterprises have are actually repetitive and mundane. I mean even before the Covid-19 situation, we have analyzed across all our IT support bots, we’re seeing that password reset accounts to about 30% of support volume for IT helpdesk applications. 

We are seeing even in customer support, if you take insurance company claim status queries, probably it accounts for about 5% to 10% of the regular volume that insurance companies get in terms of their responses. Fortunately, over a period of the last three years, the technology has adopted to be able to handle these regular and mundane customer support activities or employee support activities at a high degree of accuracy and also a high degree of CSAT and NPS.

Not only it’s automation helping improve the efficiency part, at the same time the virtual assistants are actually improving the customer experience.  Earlier the users have to wait for five minutes, 10 minutes or 20 minutes in trying to get their questions answered whereas conversational AI is helping them get those answers in a few seconds because it is able to orchestrate across these applications and users and be able to respond to the customers in real-time. 

Low Hanging Automation Use-Cases

So it’s not doing trade-offs between the customer experience that an enterprise is providing. While there are innumerable amount of scenarios that are possible using conversational AI and intelligent automation, what we wanted to bring out in this webinar is something all of us can take back and implement in the next one week. Time is really critical out here because the situation is evolving & escalating every single day. 

So three broad areas that we are seeing that are of immediate impact is primarily the customer experience. These are the times when companies that provide great customer experience even with the disruption gain a lot of customer confidence. The solution here integrates with your existing digital call center with the most popular channels like WhatsApp.  We are seeing the adoption of WhatsApp take off a big-time during this scenario where customers are expecting that they reach out to the organization and get support immediately. Lenskart is one example. 

We are seeing the Bajaj group of companies implementing this and getting significant volumes addressed right here. Customer experiences is one single piece where there is a lot of data available of your company on questions answered to your customers and that can be repurposed to automate the support on these channels. We do understand that a lot of that can be adopted through the agents where the body is not able to respond in the platform provides the ability to use the workforce that you have which is on low volume to address the really emotional the really tough customer challenges whereas most of the supporting task is automated.

Similarly HR is one function that is getting affected. People have questions on when the office is going to reopen. It’s my job going to get affected. What is the policy? So there’s a lot of uncertainty and that is driving a lot of inquiries to that HR. A lot of the answers to those queries are actually similar groups of people.  We are seeing companies ramp up adoption of HR systems during this process to provide a seamless experience to their employees, build confidence and actually deliver productivity as well happy employees drive more productivity. 

The third aspect again which is a really high impact and immediately addressable is ITSM. The environment for ideas has changed. People are working from home. They’re using VPNs more often than they used to do earlier. They don’t have an IT support guy, they can just run in and ask hey can you just reset my password, etc.

Any delay in ITSM is actually making the thin workforce that’s already loaded with volume even more unproductive because they’re not able to work on systems and applications. So these are the three broad areas that we clearly see that you could go ahead and implement immediately. 

So the call center, we talked about how that’s being disrupted. Agents are working from home. There is an increase in volume and the routing algorithms that were designed for working office kind of requirements and now are no more working as people are working from home and few of them are not even available to work.

So these are the challenges that the customer support organization is really facing and conversational AI across multiple platforms right from Google Assistant, Alexa, WhatsApp is helping them seamlessly convert their call center into digital call center within seven days and automate up to 60 to 70% and at the same time enable remote employees to seamlessly transfer from bot to human and provide delightful customer experience without missing a beat.

Conversational AI for Process Automation of HR

HR and IT automation are facing significant disruption. A lot of employee queries and conversational AI platforms with pre-built integrations with a lot of the existing HR systems, compensation systems, payroll systems. Even with the ability to integrate with a lot of policies, FAQs within the company can provide very rich employee experience while they are working from home and their regular routine is completely disturbed. 

We are seeing that channels like Slack & Microsoft Teams are being adopted and in some cases there are organizations that have enabled employee assistance on Google Assistant as well. So seamlessly employees can get their queries addressed and this communication from the HR system to engage, find out about the mood of the employees, find out about their engagement levels problems, using interactive conversations across these channels.

Diving deep into the IT process, what we are seeing across customers is there are a lot of alerts, a lot of companies are having off-premise systems that they are using. Currently our systems are not really designed to kind of seamlessly be accessible from through a work from home environment. We are seeing company employees having a lot of difficulties around those applications’ ability to access that data and the ability to use those applications in a remote scenario. 

We are seeing IT support conversational assistants that can interact and be deployed on top of your on-premise applications and enable and expose that data back to the end-user through a conversational interface.

So you take an ERP system that is within your premises, a conversational system that is integrated and deployed on top of ERP where pre-existing integrations exist and which is enabling your field sales or which is enabling your employees working from remotely to access and read and write data out of these ERP systems, sort of your core banking systems or insurance systems, is kind of driving the productivity of the employees’ multiple notches above while working from home and actually making possible some of the scenarios that were previously not possible to support in this kind of disruptions.

So the efficient orchestration of the processes, not only of the processes that are actually on the cloud but also the processes that are currently on your on-premise infrastructure but you’re not able to expose the data because they are not available remotely. Conversational AI is helping bridge that gap during this situation.

Efficient Orchestration of Processes

So helping drive efficient orchestration of processes using conversational AI. We as the company really believe that we are in it for the long game. We want to help enterprises be more effective, to help enterprises in the time of need. So we have come up with a business continuity process offered to help enterprises to a very quick implementation of a bot within seven days. We are giving this for free for the first set of processes that are more impactful for the company. We hope that will play a small role in helping you better deal with business continuity for your organization. 

So the virtual assistants are the new agents and maybe this is the start of an era where disruption is causing a positive change for the company over a period of time by enabling or by accelerating the automation within the company.

We are helping enterprises redefine the way they converse and drive business continuity

At this point in time, I’ll also like to take an example of a call center that we are working with and we’ve been working with them for the last six to seven months. These seven months were very tough for us to go ahead and replace the existing call center setup because there were a lot of existing processes that are delaying the adoption of conversational AI. So what happened over the last one or two weeks is in anyways the existing process is completely disrupted and now the company is taking this opportunity to move a thousand member call center set up to a completely digital channel within a week’s time which would be unthinkable in a regular process.

So there are areas where a significant step-change in the automation of the company is happening, taking this disruption as an advantage, as a change that can get long-term value for the company. So not only we need to look at this as a one-time business process continually but a business process continuity that is taken at this point in time and adapted to actually deliver a really long term value for the company. Some of the key features of these agents we’ve seen across they deliver about 89% of accuracy, organizations especially GCC’s are spread across multiple countries. 

So the virtual agents can handle about 92 languages so you really need not worry about deploying at multiple locations and the ability to work across 35 plus channels. So if you are using zoom or slack, you can deploy it without missing a beat across these channels. 

At the same time a big worry for a lot of organizations is that hey I cannot completely rely on this. What if the customer has something that they need to talk to the user? So all these agents have the ability to transfer it back to the human so the customer essentially gets the best of automation and the best of humans during these times.

Case Study – Customer Experience

We just run through two case studies:

Bajaj Finserv: If you just go through their site, just open the site. You would see that the virtual assistants are the first point of entry to the company and it’s configured across nine different languages. There are 100 plus bots that are deployed across different functions of the company. Not only they are automating but they are able to generate about 700 million dollars projected business through the chatbots and virtual assistants. 

From a productivity experience perspective some of the leading companies where Schlumberger have adopted our solution and using it to derive significant practical value by deploying across multiple different processes and functions. 

These are a few other examples of our deployments where customers are seeing a benefit of conversational AI automation at a really really large skill. All these applications come with pre-built integrations to your core systems, be it your cloud-based systems or be it your on-premise setups it doesn’t matter the virtual assistants can integrate and take data from that to drive those conversations and these pre-built integrations enable you to deploy within a really rapid time to deliver value.

We partner with quite a lot of software companies and are very happy to work with you if there is an existing relationship that one of your companies has with these companies and we could use those integrations relationships to deliver a great experience.

So thank you everyone and we hope we can make AI work for you. For any questions you can contact us and we’ll be able to have our consultants work with you to determine how fast they can be a bot or a virtual assistant that can be deployed. We look forward to working with you and enabling a lot of your businesses to provide great service even during the time of the disruption. I’ll be happy to take questions and go deeper as we go along. Thanks

Questions & Answers on Business Continuity Planning

Questions & Answers

What kind of NLP technologies are used?

Raghu: So the NLP technologies that we use is the core of it is Sentence Embedding. The NLP platform is built on BERT. It uses quite a bit of TensorFlow while these are the core components, the NLP will also for a lot of enterprise scenarios needs the ability to manage the content so recognition of entities through the portfolio of products a recognition of entities through semantics. So broadly it’s built on TensorFlow sentence embedding Bert as key components but the framework also involves components that are essentially tuned to how enterprise content is adapted.

How is the security aspect handled while accessing the confidential data over a chatbot?

Raghu: Masking of sensitive data. During the configuration or development of a
virtual assistant, certain entities can be configured as masked entities. All the data
that it is represented as sensitive is completely masked from users to be able to access them. Except for the APIs, no human can really access that particular data. We follow all the guidelines of ISO 27001 and that confirms a lot of security standards that are determined for the companies. Yellow Messenger is also GDPR compliant, which takes care of all the personal data that is represented. Add on to that there is an ability to time-based purging, so the user or company can set a particular timeline when the data is purged. You can say that for every half an hour all the entity level data is automatically purged and pushed back to the customers on-premise servers. That is something that can be configured.

How a human touch is added to assistants which are the core job of an HR function and its example.

Raghu: Two Examples
1. HR Solutions – We are seeing is around at this point in time is around leaves and applying for sick leave, getting their leave status, and also getting to know about work from
home policies. So a simple example is I want to apply for a leave, it goes back to the system it checks, understands what the user is asking. It can go back to the workday, success factor or surveillance system, try to get the remaining leaves for that employee and provide those leaves. Ask the minutes it’s a casual leave or sick leave actually complete the entire transaction over voice or chat.
2. Employee Engagement: the second on the engagement level which is being used phenomenally well by Flipkart, as they wanted to gauge the engagement of their users over a broad set of time and ability of the bot to proactively start conversations with employees. So to give an example during this time every evening or every two days the bot can interact with each one of the employees saying that hey it’s been already about two days, how are you feeling about productivity. Are you feeling anxious about the existing scenario and take all this feedback from the employees and give a detailed analysis to the HR team, on this is the overall mood and here are your risk areas. What do you think are not being addressed properly? 

Is a trial version available along with integration?

Raghu:  Yes, there is a trial version available along with channel integrations

Are these bots location specific?

Raghu: No. Bots are a representation of the data and content to which they are integrated or on which they are built. So if the content and data are location-specific, absolutely. But bots essentially can morph into different scenarios. It’s not about the bot, but it’s more about what you make of them. So they are not location specific by default unless you want to make them.

Is it only text-based or also supports voice/vision?

Raghu: It supports voice as well vision, but the platform supports text and voice. It can also be integrated with vision AI.

What is the role of HR and how it can ensure smooth functioning in a work from home scenario?

Raghu: I’ll take an example, leading Pharma company, they had to suddenly decide that from tomorrow we will work from home and at the same time the company was going through a lot of changes in their job grades, etc and people have like a ton of queries around their performance cycle and how is my compensation going to be affected during this particular period. How are my grades going to be affected? HR automation that was done to address a significant of these queries for the HR. 
Another company that had HR specific use cases was around reimbursements. I mean, in this scenario how do I get real reimbursements, there are so many things pending, my expenses are pending, what is the way to address that. So a lot of these are areas where HR function is bombarded and all these can be automated. The second part is a proactive engagement where HR can initiate service conversations with the users based on their mood etc and be able to get a lot of insights and do proactive measures based on those insights.

Want to automate your business processes, Schedule demo with conversational ai experts. 

video-bg is a leading CX Automation Platform helping businesses to build chatbots and voicebots to automate support, sales, marketing, HR, ITSM, in 100+ languages, across 35+ channels

Latest insights

Build your first NLP powered
intelligent chatbot in under 10 clicks