CASE STUDY

How an AI Chatbot Reduced Customer Support Tickets by 87%

Expertise Delivered

Backend
QA

Industry

Accounting

Our client,

Escalon Services is a back-office services company, providing outsourced assistance in various domains that include Finance, HR, Taxes, Payroll, Benefits, Accounting, Insurance, and Recruitment. Being an INC5000 company and with a client base of 3000+ companies across 22 countries, they continue to grow at an unprecedented rate.

Business Need

Using the platform’s inbuilt dashboard, their customers could access reports pertaining to various back-office processes. 

Since these reports were limited in their coverage, the customers had to separately reach out to the customer support team at Escalon to get an overview of metrics outside of these reports. Their customers couldn’t really drill down to a custom report, and often had to reach out to CS for additional details. 

As a result, for any custom report needed, clients would have to either email it to the support team or raise service desk tickets or wait long queues on the phone until a customer service representative was assigned to them. Most of these queries were to view custom reports related to accounting/finance.

  • This usually led to delayed first response time and high turn-around time in getting the desired result (customized reports).
  • This gave way to broken customer experience resulting in dissatisfied clients and client churn.

Seeing the need for a scalable solution, Escalon partnered with Maruti Techlabs to design an AI-based chatbot that would take dynamic queries from the customers, and immediately come back with a relevant, contextual piece of information from the database.

Solution

The bot development team at Maruti Techlabs came up with an end-to-end solution to tackle this massive customer experience problem. 

  • The team at Maruti Techlabs narrowed down to the WotNot Chatbot Platform along with Dialogflow to build the AI chatbot. We leveraged the combination of WotNot and Dialogflow to provide our Escalon with secure data integration, easier bot-scaling, and a seamless real-time experience. 
     
  • We trained the AI chatbot to identify the intents and context of the user query, using Natural Language Processing. 
     
  • The chatbot was trained on over 300 different intents and 1500 utterances to be able to manage a myriad of queries. 
     
  • NLP helped the bot extract information from the queries and retrieve information from the systems. 
     
  • We also integrated the AI chatbot with Escalon’s back-end systems using APIs, so that the bot could fit seamlessly PUSH/PULL relevant data. 
     
  • In case of complex queries that the bot was not trained to answer, it would hand over the conversation to a human agent smoothly, thereby ensuring a consistent CX. 
     
  • The AI based bot was rolled out to Escalon’s customers in a phase wise manner. Based on the success of the text-based bot solution, both Escalon and Maruti Techlabs decided to extend the chatbot to Amazon’s voice-based virtual assistant, Alexa and Google Home, as well.

Experience streamlined workflows without increasing headcount.




Phone

Don't take our word for it, take theirs!

How Does It Work?

  • This AI chatbot is placed within the Escalon SaaS product dashboard. Users are able to access the bot only after authentication, so as to protect sensitive financial data.
     
  • Once authenticated, the user types in any accounting/finance-related query.
     
  • The intelligent AI solution then identifies the intent of the user’s query and fetches the rele- vant response from the Quickbooks database (Escalon’s team used Quickbooks to manage their client account data).
     
  • The users can frame their queries in multiple ways, and still be able to receive relevant responses, as the bot was built with the capability to understand and decipher multiple utterances of the same query, while maintaining context.

Technology Stack

Finance_Chatbot.png

Result

87% decrease in the daily inflow of inbound support tickets.

✅ Query resolution speed increased by 93.75%.

✅ Over 5,475,000 queries answered annually.

✅ 25,306 hours saved annually.

✅ Predictive Analytics queries to get data from multiple systems was achieved. (e.g., Do I have enough money to process payroll? ).

Our Development Process

We follow Agile, Lean, & DevOps best practices to create a superior prototype that brings your users’ ideas to fruition through collaboration & rapid execution. Our top priority is quick reaction time & accessibility.

We really want to be your extended team, so apart from the regular meetings, you can be sure that each of our team members is one phone call, email, or message away.

Our Development Process.png
Our Process

When building a chatbot, we keep dedicated business goals in mind to achieve seamless performance across channels along with optimum automation. A natural conversation flow ensures that client expectations are met for tone, tasks, implementation, and so on.

Evaluate Use Cases
Our team analyzes the applicable use case, business objectives, target users, competition, and prospective market segment, to determine the implementation framework.
Conversation Flow Design
At this stage, our conversational experience designers step in to channel the voice of the business. They work in tandem with the development team to create a procedural flow of dialogue that determines tasks, bot personality, tone, and context of the conversation.
System Integration
After the conversation flow is formalized, our team enables seamless integration with other enterprise software (CRM, HRM, Payment Processors, Analytics tools, etc.), company knowledge bases, backend system, and target chatbot platforms in order to automate workflows and optimize bot functionality.
Training
Chatbots are constantly evolving and learning. Much like humans, they can effectively self-learn under supervision through quality training data in order to fix errors, excel at personalized communication, discern user sentiment, and follow business processes with ease.
Testing
In this phase, we rerun conversational flows, integrations and business scenarios through various test cases to achieve desired bot performance and bug-free implementation. Finally, user testing as well as application testing, is done, to cap the development process.
Deployment
This is the release phase everyone is waiting for. The intelligent bot is deployed on the company’s preferred platforms and is ready to interact with users in the execution environment.
Monitor and Optimize
Once published, the bot performance is monitored by analyzing conversation logs and usage insights to fix breakage points and reduce abandoned conversations. Analytics also help measure the bot’s ROI and run A/B tests to further optimize the use case, conversation flow and customer experience.
Evaluate Use Cases
Our team analyzes the applicable use case, business objectives, target users, competition, and prospective market segment, to determine the implementation framework.
Conversation Flow Design
At this stage, our conversational experience designers step in to channel the voice of the business. They work in tandem with the development team to create a procedural flow of dialogue that determines tasks, bot personality, tone, and context of the conversation.
System Integration
After the conversation flow is formalized, our team enables seamless integration with other enterprise software (CRM, HRM, Payment Processors, Analytics tools, etc.), company knowledge bases, backend system, and target chatbot platforms in order to automate workflows and optimize bot functionality.
Training
Chatbots are constantly evolving and learning. Much like humans, they can effectively self-learn under supervision through quality training data in order to fix errors, excel at personalized communication, discern user sentiment, and follow business processes with ease.
Testing
In this phase, we rerun conversational flows, integrations and business scenarios through various test cases to achieve desired bot performance and bug-free implementation. Finally, user testing as well as application testing, is done, to cap the development process.
Deployment
This is the release phase everyone is waiting for. The intelligent bot is deployed on the company’s preferred platforms and is ready to interact with users in the execution environment.
Monitor and Optimize
Once published, the bot performance is monitored by analyzing conversation logs and usage insights to fix breakage points and reduce abandoned conversations. Analytics also help measure the bot’s ROI and run A/B tests to further optimize the use case, conversation flow and customer experience.

More social proof incase you're still on the fence

Our Clients Review
Our Happy Clients
Review Everything
Why choose Maruti Techlabs?
13+ years experience
Start as quickly as week
Recurring cost of training and benefits - $0
Risk of failure? Next to none. 4.8/5 NPS on Clutch
Certified PMs & delivery teams
Rapid deployment & on-time delivery of projects
Complete transparency
Robust communication across shared channels
Agile & lean startup methodology
Experience across 16 industries

Years of Experience

Professionals

Projects Delivered

NPS on Clutch

Services
  • Software Product Development
  • Artificial Intelligence
  • Data Engineering
  • DevOps
  • UI/UX
  • Product Strategy
Case Study
  • DelightfulHomes (Product Development)
  • Sage Data (Product Development)
  • PhotoStat (Computer Vision)
  • UKHealth (Chatbot)
  • A20 Motors (Data Analytics)
  • Acme Corporation (Product Development)
Technologies
  • React
  • Python
  • Nodejs
  • Staff Augmentation
  • IT Outsourcing
Company
  • About Us
  • WotNot
  • Careers
  • Blog
  • Contact Us
  • Privacy Policy
mtechlogo.svg
Our Offices

USA 
5900 Balcones Dr Suite 100 
Austin, TX 78731, USA

India
302, Regency Plaza,Nr. Anandnagar Crossroads,
Satellite, Ahmedabad, Gujarat- 380015

clutch_review
goodfirms_review
Social
Social
Social
Social
©2023 Maruti TechLabs Pvt Ltd . All rights reserved.

  • Software Product Development
  • Artificial Intelligence
  • Data Engineering
  • DevOps
  • UI/UX
  • Product Strategy

  • DelightfulHomes (Product Development)
  • Sage Data (Product Development)
  • PhotoStat (Computer Vision)
  • UKHealth (Chatbot)
  • A20 Motors (Data Analytics)
  • Acme Corporation (Product Development)

  • React
  • Python
  • Nodejs
  • Staff Augmentation
  • IT Outsourcing

  • About Us
  • WotNot
  • Careers
  • Blog
  • Contact Us
  • Privacy Policy

USA 
5900 Balcones Dr Suite 100 
Austin, TX 78731, USA

India
302, Regency Plaza,Nr. Anandnagar Crossroads,
Satellite, Ahmedabad, Gujarat- 380015

©2023 Maruti TechLabs Pvt Ltd . All rights reserved.