The client

The client, a historic American multinational energy and utility corporation with over 130 years of history. This industry leader generates, transmits, and distributes electricity, serving nearly two-thirds of Oregon’s commercial and industrial sector. Recognizing the transformative potential of artificial intelligence (AI), the client saw an opportunity to optimize process-driven systems. With LTIMindtree’s BlueVerse, they aimed to address critical inefficiencies impacting both employees and contractors. These inefficiencies included spending several hours of manual effort searching for relevant information and performing various other day-to-day functions.

Need for change

The energy and utility industry are undergoing a significant transformation driven by technological advancements and the need for operational efficiency. According to a Boston Consulting Group (BCG) industry report, a significant volume of employee efficiencies can be improved by leveraging AI capabilities to assist employees. This is especially crucial as companies in the sector face mounting pressures to modernize their operations amidst evolving regulatory requirements, rising energy demands, and increasing competition.

The industry is increasingly adopting artificial intelligence (AI) to automate manual tasks, streamline workflows, and unlock new avenues for growth. These solutions drive greater efficiency, enable precise data-driven decision-making, enhance customer engagement, and improve predictive maintenance. As energy markets grow more complex and customer expectations rise, this shift is crucial for companies to stay competitive, reduce operational costs, and make better use of their resources.

Challenges

The client faced multiple challenges across several areas due to inefficient and fragmented manual processes that impact operational efficiency:

Inefficient information search

Lack of timely information and repetitive processes for performing daily functions resulted in operational inefficiencies.

Difficulty accessing real-time data

Accessing real-time data from the database was time-consuming and concerned teams needed to reach out to technical teams manually to get the data. This increased the response time and decreased the number of transactions handled.

Time-consuming template and draft creations

Manual creation of various data request templates, drafts, files, etc. in the required format required a lot of human time and effort.

Slow resolution of employee queries

Manually finding out the required information would consume a lot of time, which increased the resolution time and ultimately reduced the number of transactions handled. More staff were required to answer the calls and process the requests manually, increasing the final labor cost.

Tedious, error-prone review processes

The manual review process of various claims, purchase orders, workorder designs, etc. was time-consuming and error-prone.

Sub-optimal key words identification

The manually identified key words from various legal documents would be irrelevant and less optimal that would not be helpful in building strong legal cases.

Manual campaigns were not impactful

Manually built marketing campaigns were not only time-consuming but also less effective in creating the required impact.

Prolonged, manual hiring processes

Manual processes for designing job posts, interview guides and identification of the right pay scale(s) for the new hires were time-consuming and costly.

LTIMindtree’s solution

To help the client address process inefficiencies and growing operational complexities, we implemented BlueVerse, our AI-led transformation framework powered by AWS tech stack. This encompassed the creation and deployment of multiple GenAI bots designed to automate repetitive tasks and streamline workflows. These interventions led to smoother operations and a measurable boost in efficiency across key business processes.

Bot name	Purpose
General Chat Bot	Catered to the client’s employees with two focus areas. Large language model (LLM) capabilities to help with daily tasks such as email draft creation, doc summarization, meeting notes creation, idea brainstorming, ppt creation, Excel data visualization. Document retrieval from internal sites to provide clear and accurate information for any questions that employees have such as parental leave policy, health insurance plans, utility standards, corporate governance policies, etc.
Work Design Validation Bot	Automated pole replacement job reviews using image recognition, OCR, and text extraction to detect design violations and speed up approvals.
Reimbursement Bot	Evaluated expense reports for policy compliance. Flagged non-compliant entries for review and triggered training when needed.
Human Resources Bot	Enabled users to search for HR-related information quickly and accurately.
IT Assist Bot	Pulled relevant knowledge articles from ServiceNow to resolve IT queries and empowered users with self-service options—reducing call volumes and improving resolution times.
Legal Assist Bot	Uploaded legal documents, extracted key terms, and summarized content into user-friendly formats.
Sales & Marketing Bot	Provided customer insights (e.g., EV ownership, rebate participation) and generated draft marketing campaigns for specific segments.
Outage Assistance Bot	Delivered real-time outage information by converting natural language queries into automated SQL generation.
Purchase Bill Bot	Audited purchase orders using predefined rules and policy checks, reducing manual effort and improving procurement efficiency.
Transmission Policy Bot	Supported the policy team which caters to the heads from various internal and external teams such as transmission, advanced energy delivery, power operations, settlements, Federal Energy Regulatory Commission (FERC), Department of Environment and Quality, etc. by searching for the required information from various sources such as FERC eLibrary, CAISO, etc.
Rates & Regulatory Bot	Supported the rates and regulatory authority team to generate templates for data requests and responses based on past filings. It also reviewed past filings to understand the type of questions that can be asked by the OPUC in future filings.
Talent Acquisition Bot	Aided TA specialists and hiring managers to generate job posts swiftly using BlueVerse capabilities. It also generated interview questions based on the client’s guiding principles. In addition, it provided salary recommendations for prospective employees based on a variety of factors such as level of experience, education, interview performance, etc.

Technical Architecture Diagram of GenAI Bots

The diagram below illustrates the GenAI architecture and its various services. When a user logs into a bot, they are first authenticated. Once verified, the homepage appears, allowing them to enter a prompt. This prompt is processed by large language models, which retrieve relevant information from internal (database, cloud, etc.) and external source. The logic is built in to prevent access to sensitive or confidential data, ensuring compliance and avoiding violations.

Figure 1: Technical architecture diagram of GenAI bots

Tech stack

Front-end technologies	React, JavaScript, TypeScript, Tailwind CSS, CSS
Backend technologies	AWS Services–Cloud Base: Lambda, Step Function, ECR, Glue, Sagemaker, and Amplify
Backend technologies	AWS Notification Services: SQS, SNS, SES
Backend technologies	AWS AI Services: Bedrock (Model – Claude Sonnet 3.5 & Amazon Titan), Kendra (SharePoint Data source, S3 Data Source, WebCrawler Data Source)
Backend technologies	AWS Security Services: Cognito, AI Gateway, WebSocket, Macie, Guardrail, System Manager (parameter store), Secret Manager
Database	Snowflake, Dynamodb, SQL Server
Cloud resources	Azure (application management), web app
DevOps	Jenkins, AWS CloudFormation

Business Benefits

Implementing LTIMindtree’s BlueVerse benefitted various client teams improving their operational efficiency and saving them hours of manual effort along with massive cost savings. This included:

Expedited employees research

The various bots such as chat bot, HR, Policy, etc. accelerated research and development by automating tasks, analyzing data, generating new ideas and enhancing collaboration. This increased the overall productivity and efficiency through countless automations and process optimizations.

Streamlined design review processes

Reduction in cycle time for job approval from 12 days to <1 day.
Cost reduction from $250 to $25 (potential savings of $90,000 per year).

Automated expense reviews

Automated the reviews and categorization of $7000 expenses per month.

Reduced calls to the service desk

Reduction in user calls to the service desk team led to a significant reduction in labor cost. The bot seamlessly managed an increased volume of transactions by crawling 1400 articles.

Accelerated legal processes

Reduced the time taken to identify relevant policies to build a strong case from 4 hours to <1 hour.

Greater productivity for marketing

Increased productivity by reducing the time taken to identify target audience and draft personalized campaign emails.

Faster resolution for outages

Better availability of live outage data to take further action and reduce the overall resolution time.

Improved scalability for procurement

Increased the volume of audited purchase orders from 30 per month to 1250 per month.

Massive cost savings for the rates and regulatory team

Better information, use of past decisions and awareness of past arguments are estimated to cost recovery of $0.75 to $2 million per year.
Automatic generation of word document template saved approximately 125 hours/year (2500 data requests with an average of 3 minutes per template).

Talent acquisition savings

$200,000 saved annually by reducing the candidate hiring process to three days.

Conclusion

Powered by BlueVerse, our solution fast-tracked the client’s R&D efforts, automating repetitive tasks, uncovering insights, sparking innovation, and enabling seamless collaboration. By minimizing manual dependencies, the client achieved significant cost savings, streamlined operations, and optimized resource use.

Hyper-personalized experiences enhanced customer engagement, while AI-generated insights supported faster, more confident decision-making.

In short, GenAI became a force multiplier, elevating productivity, unlocking efficiency, and setting the foundation for scalable, sustainable success in an increasingly digital and dynamic world.

Testimonial

“Appreciate the great work that this team is implementing and thanks for the quick turnaround!

Excellent job Team for working together to put all the measures in place and for taking the ideation to the next level. Due to your tiring efforts our chat and other bots are getting cyber security approval. This was no easy task, thank you for your persistence and patience.

And hearty congratulations on accomplishing this significant milestone! Many thanks to all involved!”

– Director Division Chief Information Officer

Enhance your firm’s operational efficiency with BlueVerse capabilities; save time, enhance decisions, and transform services today.

Reach out to us at eugene.comms@ltimindtree.com

Contact us