AnnouncementIntroducing MongoDB 8.0, the fastest MongoDB ever! Read more >>

With MongDB Atlas, IGT Solutions Speeds Gen AI Time to Market

A person walking with a trolley, looking at her mobile phone.

INDUSTRY

IT services

PRODUCTS

MongoDB Atlas
MongoDB Vector Search

USE CASE

Artificial Intelligence

CUSTOMER SINCE

INTRODUCTION

A future-facing innovator

India’s economy is synonymous with business process outsourcing (BPO) and technology services, so standing out from the crowd with a truly innovative and future-facing offering takes something special. For IGT Solutions, however, innovation is the norm. Based in New Delhi, IGT Solutions operates 30 delivery centers that provide end-to-end platforms for over 25,000 users across a growing range of businesses and industries.

Many of IGT’s clients are leaders in the travel and hospitality sector—airlines, travel operators, industry bodies, and service and facilities providers. Outsourced services offered by IGT Solutions include bookings and reservations for travel companies, billing and claims processing for medical operators, and customer experience solutions like voice, email, and chat communications. For example, based on Techbud.AI, a comprehensive generative AI platform, the company offers a number of specialized AI-powered solutions for the travel and hospitality industry. These automated case-specific tools cover the full journey lifecycle, from inspiration and planning, to itineraries, baggage handling, customer requirements, and feedback.

“We’ve been in this space for two and a half decades,” says Love Ojha, Vice President at IGT Solutions. "That’s a pretty significant time, and our team now has very strong functional knowledge of how things operate.”

While none of IGT Solutions’ offerings are identical, and its client base is continually diversifying, MongoDB Atlas underpins many of the company’s solutions. Whether it’s driving innovation and business transformation with digital technology or harnessing IGT Solutions’ generative AI (gen AI) platform, TechBud.AI, MongoDB Atlas is the developer data platform helping power the innovation.

THE CHALLENGE

Rising volumes and demand

Prior to adopting MongoDB, however, IGT faced a challenge. The legacy platform offered by IGT Solutions was built on a relational databases and vector databases so relied on labor intensive workflows, involving extensive manual searching and processing of documents to answer queries or to complete specific tasks. As a result, response times were slow and operational costs high. The manual nature of the work also meant that results were prone to human error.

For IGT Solutions to continue to build its competitive advantage, the company needed to find a quicker, more functional and flexible option. IGT needed a platform that was able to:

  • Store substantial volumes of data
  • Simplify its architecture and management tasks
  • Handle both operational and vector data
  • Support large language models (LLMs), to meet the rising demand for its growing gen AI offering
  • Work across multiple cloud providers

“We’re not usually proponents of any specific technology. We don’t go to a customer and say, ‘hey, use this or use that’. We aim to use solutions that truly give the best results," says Love Ojha s.

One project for a global travel industry body gave IGT a specific set of requirements and led the business to find that there was only one viable platform to meet the demands of the client—MongoDB Atlas.

“We saw that we needed to base our solution around a document database,” explained Ashutosh Moondra General Manager at IGT Solutions. “We needed flexibility, scalability and high performance, as well as something that worked for us in terms of cost, writing style, and usage. That was the project where we first used MongoDB.”

THE SOLUTION

A complete data package

The pilot project went on to handle 50 gigabytes of data and analyze an average of 7 million tickets a day. It was also a clear demonstration of how a MongoDB Atlas Vector Search could seamlessly and securely build intelligent applications over any type of data using full-featured vector database capabilities. This would deliver faster and more reliable results than competing solutions, while ensuring data security and consistency.

“MongoDB’s flexibility gives us the convenience of being able to write and change schemas in-house as we need to,” Ashutosh Moondra notes. “Importantly, it’s a multi-cloud solution, so it doesn’t matter if a client wants to run in Azure or AWS, or even switch from one to another. It’s easy to convince them that MongoDB is the way to go.”

For IGT Solutions, this combination of factors meant MongoDB Atlas was also a highly cost-effective and compliant choice that worked both for its developers and clients.

“Having an operational database and a vector database on a single platform was another important consideration,” Moondra notes. “MongoDB Atlas Vector Search makes service integration much easier. For example, we can collect operational logs while also running our own reporting and alerting systems and maintaining compliance and security.”

“MongoDB is also reducing complexity and normalizing data for developers because it’s storing both structured and unstructured data,” adds Ashutosh Moondra, General Manager at IGT Solutions. “For gen AI purposes, we have been using Azure with MongoDB and Python at the back end. With both vector and operational data storage, MongoDB is the complete package.”

THE RESULTS

Demonstrating the power of generative AI

The results tell much of the story. Reduced development and system administration times enabled by MongoDB have led to a 20% rise in productivity and a 25% improvement in time to market. IGT Solutions’ growing range of AI Solutions built on MongoDB Atlas, are also delivering impressive improvements in performance and efficiency for end users. For example, the company’s Procurement.AI led to a 90% reduction in response times for finance queries at one client.

Moondra adds that IGT Solutions is also using its Recruitment.AIfor its own recruitment processes. This uses LLM tools to analyze thousands of resumés from a range of sources, and matches keywords against job profiles to assess applicants’ suitability.

“It will even check calendars and schedule interviews,” he adds. “The complete process is completely AI-driven, and the LLMs mean we’ve easily been able to roll it out beyond India in Spain and the Philippines. To date, it’s handled around 45,000 resumés, and the savings in time and effort have led to significant cost reductions.”

Key to the successes of IGT Solutions is the flexibility of MongoDB Atlas to securely host both operational and vector data in a single platform, and its availability on all three major cloud providers. Integrations with multiple LLMs, integration partners and embedding models also allow IGT Solutions to meet changing customer requirements. New products can be built quickly and easily, and with very few code changes.

Importantly, with MongoDB as a secure back-end foundation, TechBud.AI is now giving users across the world hands-on experience and insights into the power of gen AI.

“Some clients may have run gen AI proofs of concept for a handful of documents or files, so it’s important to demonstrate that we can scale to cover a much larger—and growing—data volume,” Ojha adds. “With IGT Solutions and MongoDB, clients can test out different AI services, LLMs and use cases without having to worry about switching between platforms or databases.”

The growing breadth of applications means IGT Solutions can now look to expand its market range. It has already added customers in new industry verticals such as mining to its roster.

“The use cases for MongoDB and TechBud.AI are expanding all the time,” Love Ojha concludes. “They’ve evolved from being simply parts of a PoC to becoming an entire end-to-end intelligence suite that can cover any function or industry.”

“With IGT Solutions and MongoDB, clients can test out different AI services, LLMs and use cases without having to worry about switching between platforms or databases.”

Love Ojha, Vice President, IGT Solutions

“For GenAI purposes, we have been using Azure with MongoDB and Python at the back end. With both vector and operational data storage, MongoDB is the complete package.”

Ashutosh Moondra, General Manager, IGT Solutions

What will your story be?

MongoDB will help you find the best solution.