Contact us
Search

  • LTIMindtree Euclid – Agentic AI for Insurance

    LTIMindtree Euclid – Agentic AI for Insurance

Insurance Tech Office

The Insurance Technology office is a dynamic hub of innovation, dedicated to transforming the insurance industry through advanced technological solutions. This team is at the cutting edge of technology, consistently developing state-of-the-art tools and platforms that cater specifically to the needs of insurance companies.

By harnessing the power of Generative AI, they have been trailblazers in creating business solutions that not only streamline operations but also enhance decision-making processes. Their platforms are designed to improve efficiency, reduce costs, and provide superior customer experiences.

The team’s pioneering efforts have positioned them as leaders in the field, driving the adoption of new technologies and setting new standards for the industry. Their work ensures that insurance companies can stay competitive in an ever-evolving market, offering innovative solutions that meet the demands of the modern world.

LTIMindtree Euclid for Insurance

Generative AI has generated significant excitement for its ability to create intelligent, multimodal responses to natural language requests. However, its limitations become evident when tackling complex business challenges requiring autonomous intelligence tailored to specific contexts. These challenges include the need for advanced data integration, deeper contextual understanding, and the mitigation of hallucination and bias-related issues. To fully harness the potential of next-generation AI capabilities, a well-designed AI ecosystem is essential—one that integrates agentic AI capabilities with robust data, models, knowledge, responsible AI principles, security measures, and governance frameworks.

Euclid is designed to deliver Enterprise grade Agentic AI ecosystem with an objective to establish a Value Extraction framework for business enterprises.

LTIMindtree Euclid delivers actionable Intelligence for Insurance business through orchestration of autonomous AI Agents designed to perform business functions across Insurance value chain.

EUCLID Platform

EUCLID Agents

EUCLID Agents

Platform for composing new Agentic AI solutions by orchestrating an existing pool of autonomous AI Agents in few minutes. This platform is hosting an exclusive marketplace of AI Agents delivering Insurance value chain specific autonomous functions. This platform will also enable a business user to create a new AI Agents from scratch and do needful configurations of data stores, knowledgebase, and tools & services for execution.

Agents: Planner Agents, Domain agents, Utility agents, Knowledge agents, Task agents, Language agents etc

EUCLID Knowledge

EUCLID Knowledge

Platform for automated discovery / creation of domain specific knowledge from unstructured docs (Policy documents, Loss Run Reports, UW Guidelines, Claims Adjudication Reports etc.) in the form of ontologies, taxonomies, and knowledge graphs. All the ontologies and knowledge graphs used by EUCLID Agents are retrieved from this platform. Key features include – Lifecycle management of ontologies & knowledge artifacts, ontologies & knowledge graphs matching and comparison views, Security & PII attributes identification etc.

Auto ontologies, homeowner ontologies, underwriting knowledge graphs, commercial insurance ontologies, claims process ontologies etc.

EUCLID Foundation

EUCLID Foundation

Core Platform for delivering enterprise ready GenAI & Agentic AI ecosystem. Key features include – GenAI Façade, RAG Pipelines, Mindful AI, Security Guardrails, Semantic Caching , FinOps, EUCLID Platform Administration and Governance etc.

In addition to Euclid foundation platform governance, it has the ability to integrate with any third-party AI governance tools

EUCLID Data

EUCLID Data

Platform offering automated data pipelines for extracting domain contextual data sets from structured and unstructured data sources. Key features include – Intelligent Document Processing, Multi-modal content extraction, Embeddings generation and vector store creation, domain contextual Data APIs for EUCLID Agents and Apps consumption etc.

EUCLID Models

EUCLID Models

Platform for automated Models / LLMs lifecycle management. Key features include – Automated collection of user’s feedback, Data exploration and readiness checks, Model fine-tuning and customization, Reinforcement Learning, Model serving, Model setup, hosting, governance and lifecycle management.

Euclid @Scale:

EUCLID @Scale

Euclid platform supports AWS, Azure and Nvidia native deployments. AWS and Azure native deployments are supported on K8s clusters. Euclid deployment at scale leverages Nvidia enterprise AI ecosystem.

LTIMindtree Euclid Microlab

LTIMindtree Microlab offers a comprehensive suite of ready-to-use GenAI capabilities in a box in a sandbox environment. Designed to accelerate your AI journey. With an exclusive GenAI sandbox available on Azure and AWS Cloud, the enterprise can explore demonstrable GenAI solutions using real datasets. The platform assists in identifying the right LLM for specific use cases and implements industry-popular GenAI solution patterns. Ensuring security and privacy, it features access controls and audit logs. Powered by Canvas AI and integrated with multiple LLMs, it allows users to bring their own data and tear down at the end of experimentation. Additionally, LTIMindtree Microlab provides consulting and handholding for both technical and functional needs, along with GenAI evaluation capabilities, including RAG and model evaluation.

Key Features:

  • Ready-to-use Gen AI Capabilities in a box
  • Exclusive GenAI sandbox on Azure/AWS Cloud
  • Demonstrable GenAI solutions using real datasets
  • Assistance in identifying the right LLM for specific use cases
  • Implementation of industry-popular GenAI solution patterns
  • Secure and private with access controls and audit logs
  • Powered by Canvas AI, integrated with multiple LLMs
  • Bring your own data and tear down at the end of experimentation
  • Consulting and handholding for tech and functional needs
  • GenAI evaluation capabilities, including RAG and model evaluation

Success Stories

Financial Research Co-Pilot

Increased Productivity & Accuracy with 60% Savings

Business Problem

  • Fund Research Experts spend between 40-80 hours per report. These reports are consumed by enterprises who are seeking to maximize their investment
  • Most of the research work is manual and customers may not be willing to wait for two weeks for such reports

Business Solution

  • Boost the productivity of the financial researcher by automating the drafting of the financial research reports using Gen AI’s Natural Language Generation capabilities
  • Pre-train the Gen AI models using the previous financial research reports to reflect the human writing style compared to machine-generated reports
  • Leverage Gen AI capabilities for summarizing large source PDFs and generating relevant research report sections

GenAI Service Used

  • Azure OpenAI GPT 3.5, AWS S3, Python 3.11

AI-Powered Claim Summarization for RMs

Boosts Efficiency by automating claim analysis, reducing manual work, and improving decision-making accuracy.

Business Problem

  • Currently, Resolution Managers (RMs) are required to go through multiple sections and pages of the Claim Notebook application to obtain key information and status of a claim.
  • This process can be time-consuming and tedious, as users must carefully analyze each section to piece together the overall status of the claim. This includes cross-referencing information from different sections to ensure accuracy and completeness.
  • This manual process can be particularly challenging for complex and/or lengthy claims.

Business Solution

  • Claim Summarizer tool is intended to provide RMs with a way to automatically compile information across sections and pages of Claim Notebook, helping the RM to quickly understand the status and most critical aspects of a claim.
  • The Claim Summarizer is intended to be an aide (“co-pilot”) to the RM; when using the Claim Summarizer report, RMs should follow up appropriately to confirm accuracy and manage claims appropriately.

GenAI Service Used

  • Llama Index and Azure OpenAI GPT-4o.

Streamlining Underwriting Workflows with AI-Powered Solutions

Streamlined Unstructured Data Processing Boosted Productivity and Reduced Time Consumption

Business Problem

  • BSI Agents send emails to UWs who then need to get information from emails, attachments, forms and create an account. After account creation, they have to extract additional information to create a submission.
  • Approx 270 emails per week for AI Intake – for 315 users with Average time taken 2 hours to 2 days to process the request manually.
  • This manual process is too effort-intensive and takes a long time, which keeps UWs away from their core work.
  • This also causes to skip some opportunities as Underwriters Bandwidth is taken.

Business Solution

  • Emails sent directly to a generic mailbox, then to S3 bucket.
  • Text Extract models are used to extract information from emails and attachments to integrate into automated account creation as well as submission.
  • The average time to process each email was reduced from approximately 4 hours to just 15 minutes. This 15-minute window applies to exception cases where there are failures or when additional information beyond standard data is required.
  • Standard emails are processed automatically and queued in the work queue for underwriters, minimizing delays.

Ask Anything Chatbot for Customers

30-40% Productivity Gain, 2X First Call Resolutions, Improved CSAT Scores, 50% Faster Training Time.

Business Problem

  • Customers often face difficulties accessing information about their insurance policies, such as Coverage details, Claim submission processes & Policy expiration dates.

Business Solution

Ask Anything- Q&A Capability

  • Chat Style search across Enterprise Docs
  • Integration with Customer and Policy Data which is further extensible to end-customers for Self-Service

GenAI Service Used

  • Add GenAI Service Used

Digital Transformation: Case Summarizer

20-30% Gain in Policy Issuance Time, 20-25% Productivity Gain, Streamline Case Resolution, Faster Issue Resolutions

Business Problem

  • Difficulty in tracking case details such as when records were received and who the last reminder was sent to, Long policy issuance times and service issue resolutions.

Business Solution

Summarize Case information and provide Q&A functionality for given cases for efficient tracking of case details.

  • Summarize new business and inforce requests
  • Customized Summary Creation on Docs- Forms, License
  • Real-Time assistance and Next best actions to issues

GenAI Service Used

  • Add GenAI Service Used

AI-Driven Captioning and Semantic Matching for Enhanced Asset Retrieval

Streamlined Captioning and Image Matching Boosted Newsletter Efficiency by 60-70%

Business Problem

  • The client was managing a system where images, videos, and documents were tagged based on result statements. This manual tagging process was inefficient and inaccurate. Every time new result statements were added, the tags had to be updated manually.
  • Additionally, the system struggled to effectively match relevant assets (like images and videos) to these statements, leading to poor content retrieval and a lack of useful results.

Business Solution

  • The proposed solution involved leveraging AI to automate the tagging process. Instead of manually assigning tags, an AI-driven model was introduced to automatically generate captions for images, videos, and documents. The solution includes:
  • Image Captioning: Extracting descriptions for images, up to 60 words, using AI.
  • Video Captioning: Extracting descriptions from video frames and summarizing background audio using AI, converting audio to text, translating to English, and combining it with the generated visual descriptions.
  • Document Processing: Summarizing text from word documents to generate concise descriptions.
  • NLP-based Matching: Using a sentence similarity model to compute matching percentages between assets and result statements based on semantic meaning.

GenAI Service Used

  • GPT Vision-4 Preview

Requirement Bot: Automated Product Filing for SMEs

Achieved a significant reduction in the time for product filing across different coverages and attributes with an overall accuracy above 90% in product filing.

Business Problem

  • PSIC BA/ SME spending manual efforts in going through the Product Filings, Rate files and JIRA/ADO, etc to produce a template called “Product definition”, which is time consuming.

Business Solution

  • An Application Interface to take in a Product Filing as an Input and generate an output of Requirement Matrix(es) Stored the insurance policy documents shared by the client in AWS S3, the embeddings for these document were stored in vector DB, AWS Aurora.
  • Implemented RAG (Retrieval-Augmented Generation) to extract information from the insurance documents.
  • Leveraged Prompt Engineering (prompt tuning, few-shots learning, etc.), Parallel/Multi pre-processing techniques to customize Large Language Models for optimal performance for different coverages.
  • Compared different LLMs with rouge score to select the best performing model for this use case. Generate the final excel file using a fully automated mechanism & a UI to integrate with the solution.

GenAI Service Used

  • Claude Sonnet, Claude V2.1, Jurassic 2 Ultra, Titan Embeddings G1 – Text

Centralizing Knowledge Database for Enhanced Self-Service

Access to Information Boosted Efficiency and User Autonomy through a Centralized Knowledge Base and Chatbot Integration

Business Problem

  • Utilize disparate organization data sources to create a central Knowledge Database to be used for Self Service
  • Aspen Business mentioned that Aspen IT is slow, it is less responsive, it is not having state of the art systems.

Business Solution

  • Create a Chat BOT (Similar to Ultima BOT) and integrate with MS-Teams to provide a self-service DB solution for user across the organization.
  • By implementing self-service for password reset requests, we can achieve a 20-25% efficiency gain. Currently, we handle 500-600 requests per month, with about 25% (125-150) being password resets.
  • With self-service, we can reduce this by approximately 40 tickets. Additionally, this improvement is expected to enhance the end-user experience, increasing our Customer Satisfaction (CSAT) score from the current 3/5 to 4/5.

GenAI Service Used

  • React | Angular | Python | Java | Mongo DB | Chroma DB (Vector Store) | Airflow (Orchestration) | Kafka (Middleware)

Boosting Migration to Achieve Efficiency for IT Team

25% Faster time to market & Boosted Code Conversion Efficiency from 20% to 70%

Business Problem

  • Migrate the code from one stack to another. for Ex: Stored Procedures to Rest API & Java to Springboot (upgrading from Legacy to Latest stack), to reduce both CAPEX and OPEX.
  • Cobol Documentation (Legacy Mainframe to CloudBase) to reduce both CAPEX and OPEX)
  • pyUnit/jUnit Test Case Generation for better Engineering Efficiency

Business Solution

  • Implement appropriate Prompt Engineering before feeding the code for conversion. Utilize AWS Claude to Enhance Engineering Efficiency
  • Splitting the Code having more than 200K tokens into multiple chunks
  • Efficiency of the code conversion improved 20% to 70% efficiency based on the Use Case compared to Manual Effort

GenAI Service Used

  • AWS CLAUDE, BEDROCK

Accelerating Software Development using GitHub Copilot for Unit Testing

Significantly enhanced development process by reducing Time-to-Market by 30%, Lower Defect Leakage rate

Business Problem

  • User Story Linkages are managed in rally currently but when a particular feature is changed, or an enhancement request comes in it need to be ensured that the entire test suite is generated with all the linkages and with all scenarios which can be covered by a unit test.

Business Solution

  • Unit Testing Coverage for Shortened Testing Cycles to improve time to market.
  • Using GitHub Copilot for Unit Testing

GenAI Service Used

  • Github Copilot

Reimagining Legow Application with CodeBrew’s Gen AI

Legacy modernization project completed in Approx 40% lesser time than the actual estimated time duration, i.e. From 4 months to 1.5 months

Business Problem

  • The Legow application was designed to work exclusively with Windows 2008 operating system which is currently outdated. It was impossible to migrate to newer windows versions such as Windows 2019 or 2022. As a result, moving away from the legacy Windows 2008 environment was not feasible. Moreover, attempting to reverse engineer or improve the application was made difficult by the absence of the source code.

Business Solution

  • Leveraging CodeBrew tool of Insurance GenAI Marketplace for improved developer productivity, the source code was rewritten to replicate the existing functionalities without introducing any new feature or enhancements.

GenAI Service Used

  • .Net MVC, Oracle 19C to rewrite Java/jsp & Oracle 9i, GenAI

Contact Us