POV

Intelligent Automation
- A Boon for the Insurance Industry
 Intelligent Automation - A Boon for the Insurance Industry

Introduction
Despite modernization, some insurance companies dealing with life and General insurance are still
stuck in time-consuming back-office processes. This involves huge volumes of repetitive business
operations and monotonous clerical tasks that often pose a challenge in fulfilling growth targets
and delivering quality customer service. Such tasks also add to operational costs, further limiting the
pace of growth, and resulting in sub-par products & services, making it hard to maintain a strategic
competitive advantage in the Insurance market.

One of the key tenets of Intelligent Automation (IA) is Robotic Process Automation (RPA). Being
reported as the fastest growing segment of the global enterprise software market in 2018 by Gartner,
RPA is already helping insurers build a high-growth responsive business while optimizing costs.
Software bots emulate transactional, administrative tasks that are repetitive, rules-based and require
minimal decision making or strategizing. When RPA is complemented with other cutting-edge digital
technologies such as Advanced Optical Character Recognition (OCR), Machine Learning, chatbots
and cognitive technologies, it goes beyond basic automation by learning and evolves as it automates,
thereby delivering exponential value.

Transforming Insurance with

Intelligent Automation
Intelligent Automation is already changing the way Insurance companies operate. The industrious
workforce of software bots not only emulates repetitive and rule-based transactional and administrative
tasks, but also takes care of certain levels of elementary decision-making. Human domain experts
are then free to focus on more strategic tasks that involve reasoning, judgment, and emotional
intelligence. This digital workforce of bots is getting smarter with Cognitive Automation enabled by
transformational technologies like Machine Learning (ML), Computer Vision, Artificial Intelligence (AI),
and Natural Language Processing (NLP). The typical areas where Intelligent Automation is making a
huge difference are:

Policy Administration and Servicing Renewal Processing

Claims Processing Underwriting
Regulatory Compliance Process and Business Analytics
Sales and Distribution Finance and Accounts

2
 Intelligent Automation - A Boon for the Insurance Industry

Some key functions that are being automated by Intelligent Automation include:

Data entry into various systems including web portals, desktop applications and legacy systems

Data extraction from various document formats such as Word, Excel, PDF documents, images
and emails

Data retrieval and updates to various types of databases such as MS Access, SQL, Oracle etc.

Interactions with web services and reading writing XMLs, JSON file formats

Automation on thin clients as well as thick clients

Case in point:
Insurance underwriting submissions
In the Insurance industry, the underwriting submissions intake from managing general agents (MGA)
and direct business (in-house underwriters) includes creation of new submission, its clearance and
the binding of the policy. The end-to-end submissions intake process can be automated, including
submission requests from MGA systems, via a web-service or from Direct Business through emails for
automating client creation, submissions, clearance and policy binding.

Multiple systems will typically be involved in the end-to-end process such as:

1. A third-party system sending in submissions in XML/ JSON files through a Web Service

2. SQLs/tables in a database

3. Multiple screens on a front-end application

4. Submissions mailbox (e.g. Outlook) with possibly a custom outlook data entry add-in/form

This process is complicated, long-winded, and can cost Insurance companies thousands of manhours
and operational overheads. They need a faster process that free up underwriters to focus on the core
business rather than mundane operational activities.

3
 Intelligent Automation - A Boon for the Insurance Industry

Automating MGA submissions intake

A powerful IA solution can consist of a robust framework based on the below tenets:

1. Orchestrated automation using queues and transactions

2. Unattended dispatcher and performer bots for submissions clearance and policy binding.

The MGA submissions intake automated process can be split into two parts:

Submission entry and clearance

The MGA will feed new submission information of the insurer into its portal, which will further invoke
a web service. This web service will update the submission data into the database storing the entire
JSON received.

The submission entry bot will fetch this JSON data from the database, de-serialize it and convert it
into data dictionary. Furthermore, the bot will also check if the client exists already. If it does, the bot
will perform the submission by entering this data into the insurance company’s portal and update the
generated submission number into the database. If the client does not exist, the bot will create a new
client, update the client number in the database and then perform the submission.

The Automated Clearance process will concurrently run in the underwriting system portal for the
submission entered. This process will then update the status to reflect the relevant status whether the
submission has been “Cleared”, “Declined” or a “Clearance” is requested. The submission entry bot will
update the database table (e.g. SQL database) with the Submission Status.

The MGA portal will keep checking for a clearance response by hitting the webservice with the
reference number every few seconds.

4
 Intelligent Automation - A Boon for the Insurance Industry

Auto Clearance
Webservice process is initiated
Data in Json
MGA/agent enters new in underwriting Bot fetches
format
Start submits Insured
is sent to
submission data in system and the data from
details SQL table with database table is the database
Webservice
status as ‘Received’ updated with Cleared/
Decline status

Bot checks for

existing client
Legend
Insurance company’s Web Service

Company’s underwriting system portal

BOT

MGA/ Agent’s Portal YES

Client exists?
SQL Server

NO

Bot perform new submission by Bot create a new client by

Agent’s Portal invokes entering entering the
End Webservice with reference data into the underwriting data into the underwriting
number for Clearance status system portal and generates system portal and generates
submission id client number

MGA: Submission entry and clearance

5
 Intelligent Automation - A Boon for the Insurance Industry

Policy binding
Post submission clearance, the agent will initiate the quoting and binding process in its portal and on
binding a quote - the policy information will be sent in JSON format to the company’s webservice with
the matching reference number used to identify its relevant submission ID. Webservice will then enter
this new policy entry into the database.

The Policy Entry bot will pick up the policy data from the database in JSON format, create a quote
and then bind it in the company’s insurance portal by entering the relevant data into the portal. The
bot will also update the database with the newly generated policy number and mark the status as
‘Entered’ thereafter.

MGA/ Agent’s portal sends Policy data in JSON format sent Webservice enters the new
Start the policy JSON with reference to insurance Policy details
number for binding company’s webservice in database table

Legend
Insurance company’s Web Service

Company’s underwriting system portal

Bot fetches the data from

BOT
the database
MGA/ Agent’s Portal

SQL Server

Policy Entry BOT creates the

Policy Entry BOT updates the
Agent’s Portal invokes Quote and Binds policy in
End Webservice to
database table with Policy
company’s
number,
get the new policy number underwriting system portal and
status as ‘Entered’
generates policy number

MGA: Policy binding

6
 Intelligent Automation - A Boon for the Insurance Industry

Automating direct business

The Direct Business submissions intake end-to-end process can effectively use RPA with cognitive
OCR combined with machine learning to yield incremental accuracy as the ML model learns over
time, provided there are simple to medium complexity documents with decently structured and
uniform layouts. A ‘human in the loop’ is inevitable where poorly scanned image-based documents,
unstructured emails and manuscripts come into play.

The process will start with a bot scanning underwriting email account for rule-based submissions.
The bot will then categorize the emails, extract the relevant email content along with the attached
PDF submission forms/documents and feed the data into the underwriting system. The bot can also
interact with Outlook add-ins for processing the submissions.

The direct business submission clearance process will typically involve checking the Direct Side
Submission Clearance Outlook mailbox which is part of the Insurance company’s underwriting
office for clearance emails that come in from different brokers. The mails will very likely come with
attachments in form of various file types such as PDF documents, Word files, Excel spreadsheets etc.

The cognitive bot with the help of OCR will extract the relevant information from the attachments
such as name, address etc. and store it in the database. This data will then be filled by the bot in a
predefined add-in likely to be present in the underwriter’s mail client, such as Outlook. After entering
the status of submission and reason in the add-in an automated email will be sent to the bot’s Outlook
account.

The bot will now fetch data in JSON format from the database, de-serialize it and convert it into data
dictionary. Furthermore, the bot will check if the client already exists. If it does, bot will perform the
submission by entering this data into company’s underwriting system portal and update the generated
submission number into the database. If not, bot will create a new client, update the client number in
the database and then perform the submission. The bot will also store the submission number and its
corresponding policy type in the database.

The next step would be to mark the submission corresponding its policy along with the reason. To do
this, the Insurance company could possibly have its own content capture Outlook add-in. The bot will
highlight the mail received in its mailbox by looking at a specific subject line and click on the add-in.
In the add-in, the bot will fill the required details, submission clearance status, and a reason and mark
the submission.

7
 Intelligent Automation - A Boon for the Insurance Industry

Start

Broker’s web service sent mail Cognitive BOT and/or OCR

Information is stored Bot fetch the data
to the insurance company’s extract relevant information from
in the database from the database
underwriter the attachments in the mail

Broker’s web service sent mail Cognitive BOT and/or OCR

Information is stored Bot fetch the data
to the insurance company’s extract relevant information from
in the database from the database
underwriter the attachments in the mail

NO
Bot fetch the data
Client exists?
from the database

YES

BOT adds the submission

details/product details and
generates the submission id

BOT uploads the

BOT updates the
attachments into the
The BOT reads the email submission number BOT clicks on submit
BOT highlights the email form opened from
and extract the status and status (Cleared/ button to
and clicks on company’s add-in. BOT also
(cleared/declined) and Declined) with reason upload the details into
content capture add-in enters document type
the reason as per extracted the form
corresponding to the
information
document

Legend End

Email BOT Add-In

Underwriting Database
system portal

Direct Business: Submissions

8
 Intelligent Automation - A Boon for the Insurance Industry

A successful automation of both MGA and Direct Business submissions intake processes could yield
the following benefits significantly that can be tangible if measured with the right sets of data:

Efficient handling an average load of thousands of yearly submissions

Reduction in transaction Turnaround Time (TAT)

Reduction in human efforts

Reduction in Average Handling Time (AHT)

High first-pass yield accuracy interacting with different systems (excluding document extraction)

Eliminate errors that would typically be caused by humans

Challenges in implementing IA
Staying in the comfort zone

Several Insurance organizations have not been able to aggressively run successful automation
initiatives due to resistance from within the company and an aversion to change. Automation often
tends to be considered an overhead, and a lack of synergy between operations and IT further adds to
such doubts, thereby lowering the confidence to kickstart automation initiatives.

Optimizing ROI and prioritizing processes

Organizations need to look at designing a robust tool for calculating their ROI, which would include a
comprehensive questionnaire on process complexities, technical intricacies and calculations. An over-
zealous management could expect RPA to be a magic wand to automate any kind of processes. But,
a shallow due-diligence often leads to sub-par results. Some of the costs to be considered is the cost
of bots, licensing fees, bot monitoring, support and maintenance costs, costs incurred due to the
underlying infrastructure/ applications / system changes, etc. With the right prioritization of processes
to be automated, the ROI can be encouraging and fruitful in the long-term. To achieve the right
prioritization and ROI, an extremely vigilant due diligence is of paramount importance.

9
 Intelligent Automation - A Boon for the Insurance Industry

Legacy systems and data standardization

Often Insurance companies have a variety of systems including CRM, ERP, policy admin, claims
processing, web applications and legacy systems and that could complicate IA implementations.
With the constant churn of data in varying volumes and different formats being generated by various
systems, standardizing the data and formats is a big challenge that Insurance companies constantly
face. IA implementation will always be a challenge in a frequently changing and uncertain environment.
But the right set of tools combined with different complementary digital levers like RPA, OCR and
cognitive AI based tools should make things easy.

A tailored suit fits better

The principle of one-size fits all does not really work well with Insurance companies. This is because
each of them deals with different data sets, formats and business volumes. So, a customized IA
framework is usually required to get the job done with high consistency and reliability. Also, there are
several compliances that any Insurance firm needs to follow. Hence, it is imperative that the automation
solution allows for integration and customization based on the changing compliance expectations.

10
 Intelligent Automation - A Boon for the Insurance Industry

Opportunities to implement IA
in the underwriting lifecycle
Typical potential RPA enablement cases in the underwriting world in P&C Insurance are depicted in
the heat map below.

Submission to
Quote to Bind Bind to Book Book to Issue Mid Term Servicing
UW Life Cycle Clearance
Submission
Post Bind
Clearance & Risk Selection Rate Price Quote Policy Issuance Policy Servicing
Different Stages Triage
Coding

Commercial and Capture Follow up on missing Mid term policy

Document Risk UW Quality Premium
Specialty Lines of Exposure Quote Entry contingencies/Missin changes Premium and
Indexing Qualification Assurance Booking
Business Info g Information Non Premium bearing

Auto, Property, Cancellations,

WC, Marine, Reading Forms & Schedule Generate Declaration updates
Risk Services Rating Plans Billing Setup Reinstatements/DNOC,
Umbrella , GL, Extracting data Rating Quote and Endorsements
Short Rated
Enviro, Property,
Health, Casualty, Drafting the Policy Certificates of
Reconciliation Document, Issue the Insurance, Additional
Aviation , Political New Submission Loss Manage
Reports Referrals of Bind to Book final policy to UW Mortgage, Renewal
risk etc.. & Renewal Setup Forecasting revisions
records &Policy ageing chase Processing, Premium
Finance

Activities UW Follow up Metric and Adhoc Agency Code, Loss

Applying Blocking UW Rules & Account Issue
Rate on Bind not Reporting to Runs, Billing Plans
& Clearing rules Guidelines Pricing Proposal
booked cases stakeholders changes
Personal Lines of Follow up on Experience Mod,
Acknowledgemen UW “Go /No Go” Overrides of Notify
Business Re Rate Open cases to Manage Documents Payroll Changes,
t letters Decision Premiums Stakeholders
be booked Premium Finance
Personal Auto,
Personal Property Status Changes –
Binder /Policy Contract Paper and Electronic
NTU/Declinature Monthly Closes
Other Peripherals Preparation Management filings
& ARH

Automation Potential High Medium Low

Heatmap of Underwriting lifecycle in P&C Insurance

11
 Intelligent Automation - A Boon for the Insurance Industry

Setting up a holistic automation

ecosystem in your organization
As enterprises evolve digitally, it becomes imperative to build a holistic automation ecosystem within
the organization where multiple digital technologies can co-exist and work towards the same business
goals. E.g. RPA combined with other digital technologies like OCR, Workflow Orchestration, Chatbots
and Cognitive Automation built on machine learning models is already creating a significant uplift
in operational efficiencies across critical functions in the Insurance industry. Personal assistance on
submission of important information and guiding the customer on next steps is seamlessly handled
by chatbots along with digital interfaces.

IA backed with predictive business analytics as well as operational analytics has been successful in
accurate and fast tracking multiple repetitive Insurance processes in policy administration and claims
in turn delivering higher process returns. Advanced OCR capabilities combined with ML would not
only help extract and validate data from a plethora of documents, but also help reduce the cost,
time and effort wasted in processing these documents prone to human errors. Advanced OCR
can be used in many use cases in Insurance like submission documents, quote, policy, claim forms,
MTA, renewals, invoicing and so on. Going forward, IA will not just free humans from mundane
tasks to perform strategic & value adding roles but also work collaboratively along with their human
counterparts to handle end to end role based functions to transform business operations and customer
experiences drastically.

Saurabh Navangul
Senior Project Manager, Intelligent RPA Practice, LTI
Saurabh comes with a rich and diversified experience of 18+ years in managing multi-million-
dollar ADM, Testing & RPA engagements with Fortune 100 clients across geographies and
industries including Life and P&C Insurance. In his current role, he is responsible for powering
LTI’s Insurance customers with tightly integrated Intelligent RPA solutions, ensuring
enhanced customer experience and seamless delivery. He is a UiPath and Automation
Anywhere advanced certified professional, and also holds other certifications in Project
Management, Selenium Test Automation and Google Analytics.

12
LTI (NSE: LTI, BSE: 540005) is a global technology consulting and digital solutions company helping more than 420 clients

succeed in a converging world. With operations in 32 countries, we go the extra mile for our clients and accelerate their digital

transformation with LTI’s Mosaic platform enabling their mobile, social, analytics, IoT and cloud journeys. Founded in 1997 as a

subsidiary of Larsen & Toubro Limited, our unique heritage gives us unrivaled real-world expertise to solve the most complex

challenges of enterprises across all industries. Each day, our team of more than 30,000 LTItes enable our clients to improve the

effectiveness of their business and technology operations, and deliver value to their customers, employees and shareholders.

Find more at www.Lntinfotech.com or follow us at @LTI_Global

info@lntinfotech