Current customers face lots of pressure to support their existing customer base via the call center due to Covid-19 situation.The call center agents were forced to work from home because of the lockdown and there were only a few of them able to support their customers as many went into quarantine. Hence, a quick automation solution was needed to increase operational efficiency and reduce CSR handling time.
Cross-Bot Automation for Customer Service
Pega Hackathon 2020 | Ai4Process Ltd | Cross-Bot for Customer Service with Chatbot and Emailbot
Value Proposition/Product Description
A quick (4 week production Go-Live) Out-of-box Pega solution to meet the automation needs for customer service by reducing the CSR handling time during the peak of omni-channel demand by more advanced digital customers. This solution not only considers the customer micro-journey but also helps the employee's journey by making the employee more productive in their work by focusing on real customer problems. Automation reduces the time for both the customer & employee journey, and also keeps both happy by giving them a 360 degree view of the customer interactions . Cross channel bots are a step ahead of omni-channel and help answer customer queries fast and aids retaining them in the organizations. This solution also helps with STP (Straight Through Processing) by Text Analysis, Machine Learning, OCR, pre-filled PDF generation, case management, all built in Pega and reducing the licence cost for third party components.
The objective of this product is to tackle new issues in our evolving world in terms of humanitarianism, economic relief and disaster recovery by improving access to automated services when normal human based services are adversely affected.
Benefits For Business
- Supports remote working model for CSRs
- Reduces the CSR workload by NLP processing
- Conversation history is saved with the case instance
Benefits For Customers
- Helps to reduce the call center wait time
- Customers can access and log requests at anytime and from anywhere
What it does
- The Customer – Kristina, has just moved house.
- She goes to the AI4 Services webpage to inform them of her new address.
- In the chat window she starts a conversation and then selects the change of address option.
- The AI4 Services Pega chatbot receives the change of address message.
- The Pega Chatbot then starts the change of address service process, which asks Kristina to enter her account number and then her new address.
- The Chatbot then tells Christine to expect an email on her registered email address with further instructions to complete the process.
- The Pega automated service process then generates a PDF form pre-filled with Kristina’s details including the new address and emails this to Kristina.
- The email asks her to sign and return the form including a document as proof of her new address.
- Kristina can either fill in the form and sign it electronically (if she has the capability) or she can print, fill, sign and scan the form and then email this back to Ai4 Services with a proof of address document also attached.
- AI4 Services Pega email Bot receives Kristina’s email and uses OCR to read the scanned PDF attachment.
- Text analysis and machine learning identifies the email as a response to an existing service case so it attaches the email to the case and routes an assignment to an agent work queue.
- The CSR and managers can view a report of pending and completed Change of Address cases to manage the work load.
- When Kristina’s email has been received the service case status changes from Pending-Response to Pending-Review.
- The CSR Either picks the new case from the worklist, or receives the case when selecting the Get Most Urgent button.
- The CSR reviews the data pre-populated in the case and checks the attachments sent by Kristina are in order.
- The CSR can view the history of the interactions with the customer in the case audit and case narrative.
- If the case is not in order, the CSR can use the suggested replies to email Kristina back for any further clarification.
- The CSR authorises the change of address in the system records.
- The process then sends Christina a confirmation email and the case is resolved.
- Kristina receives the confirmation email as proof the change has been completed successfully.
- The CSR returns to the report to view the current workload ( which shows Kristina’s case has been resolved) and views the Chatbot conversation that Kristina had for background information.
Youtube channel Ai4Process consist of two video (usecase and real demo video)
Usecase Video (Animation)
Real Demo Video (Working example)
How we built it
We created micro journeys and implemented email bot first and enhanced it to chat bot along with the NLP (Ruta & ML) after this we implemented pdf generation with the pre-filled values followed by integrating with OCR component
Challenges we ran into
Configuring Abbyy to read formatted complicated forms and scanned documents. Initially the output of the OCR is un-formatted and we managed to retrieve formatted text by updating the abbyyTextExtractor parameters
Accomplishments that We're proud of
We have a generic solution which can be deployed to any customer within 4 weeks
What we learned
We Learned Pega capability around emaibots, chatbots, NLP, Machine Learning, PDF generation, OCR, etc We are now capable of advising customers on Pega automation capabilities and are less dependent on 3rd party components which reduces the cost at customer side.
What's next for Cross-Bot Automation for customer service
- Pushing Pega UI sections to the customer via chatbot
- Handing over the chat conversation to the live agent
- Integration with Docusign
- Integration with Unified Messaging component