Build a Secure Data Backbone
That Scales — Reliably

The channel marketing automation company collaborates with enterprises, accelerating channel-driven demand generation, strengthening partner engagement, and growing channel revenue.

It wanted to modernize its platform by moving from monolith to microservices architecture to improve the feature cycle time and increase user engagement.

The implementation involved the challenges below-

• Understanding existing legacy systems by analyzing codebase and architecture to learn about dependencies and potential areas for improvement
• Co-existence of legacy and new systems for live users
• Prioritizing the right features by developing a strategic roadmap guiding modernization efforts

1. Strangler fig design pattern
   • Replaced functionalities of the legacy system with new microservices or modules.
   • Ensured seamless co-existence and gradual transition from the legacy system to the new architecture.
2. Anti-Corruption layer
   • Implemented an anti-corruption layer to shield microservices from being influenced or corrupted by legacy system dependencies.
   • Defined clear interfaces and protocols to prevent contamination of new services by legacy code or data structures.
3. Micro frontend
   • Integrated a React application as an iframe within the existing .NET application.
   • Enabled independent development and deployment of frontend modules while maintaining a cohesive user experience.
4. Microservice architecture with EKS (Elastic Kubernetes Service)
   • Adopted a microservice architecture pattern leveraging EKS for container orchestration and management.
   • Enabled scalability, fault tolerance, and efficient resource allocation through Kubernetes-based infrastructure.
5. HTTPS and Queue (SQS and SNS) for Communication between Legacy and New System
   • Implemented HTTPS for secure communication between the legacy and new systems, ensuring data integrity and confidentiality.
   • Utilized Amazon Simple Queue Service (SQS) and Simple Notification Service (SNS) for reliable and asynchronous messaging between components of both systems.

• Improved development speed and go-to-market strategy
• Simplified features with better user experience.
• Accelerated customer acquisition.
• Enabled direct integration to platform – New Business Offering.

We worked with a supply chain management software platform that has operations in 20 countries with 10K buyers and 200K suppliers on the platform. Its existing solution was not suitable to handle scalability. With business expanding, the company decided to modernize the system to improve concurrency control.

• The existing solution was using an online excel to manage Bill of Material (BOM) from buyers and responses from sellers. But it could handle only 500 rows and 20 columns concurrently.
• The company had to process larger BOMs manually. It used to take 5-6 months to reward single BOMs, and they wanted to reduce it to 5-6 weeks.
• They wanted to modernize their system to handle 10K rows and 600 columns concurrently for 100 users.

• We built a frontend like excel with responsive backend APIs, which could handle a concurrency of 3K requests/second. It allowed 100 users with different roles to make changes/format the sheet simultaneously.
• To support 10K rows and 600 columns for every BOM, we migrated existing SQL to scalable NoSQL (MongoDB) and migrated .NET monolith to node.js based microservices.
• Forwarding, we created an online excel supporting over 10 million search, filtering and sorting options using our own database on Lucene search.

Scaled a platform to handle more than 10K rows and 600+ columns for 100+ concurrent users.

We teamed up with a recruitment automation platform that helps recruiters organize recruitment for fast growing lean organizations. The initial product was built 17 years ago. Since then, the client added many functionalities.

The existing on-premise version faced revenue growth challenges as many organizations opted for the pay-as-you-go model. Adding new functionality to the legacy product also consumed a lot of time and it inspired the client to think about re-architecture.

1. Architectural limitations
   • Our client was spending too much time adding new features as the process was impacting existing ones. Fixing these issues was time-consuming and affected developer morale.
   • It was built using monolith architecture and old technologies and was not supporting multiple databases. It had a lot of boilerplate code for security and session expiry, and comprehensive test cases were not written.
2. Poor user experience
   • User interactions were not intuitive. Training recruiters used to take a lot of time.
   • It was not supporting custom reporting and analytics.
3. Difficult to upgrade product versions
   • Upgrading to the latest version was difficult as multiple versions of the product were created to implement customer-specific features.

1. Migrated to a single-page application
   • The tech stack was migrated from Java 7, JSP Servlets, and SOAP services to a single-page application using Java 8 + Spring Boot.
   • Vue.js and AngularJS were used for the frontend along with a Solr-based search engine and Redis cache.
2. Migrated from monolith to a modular monolith
   • Separate modules were implemented for admin, recruitment, chatbot, authentication, and integrations with Microsoft Teams, Google Meet, and HackerRank.
3. Implemented multi-tenancy
   • Enabled to lower the cost of adoption and open new revenue streams from small and medium-sized companies.

Re-architecture

• Increased market share by penetrating small and medium-sized companies.
• Accelerated feature development by 2x.

A cloud-based media and entertainment software provider managing TV advertising inventory, radio trafficking, finance, revenue management, scheduling, rating analysis, and business intelligence needed to modernize and expand their advertising business platform.

The customer faced several key challenges:

• Legacy Systems: An outdated analytics system required replacement to enable real-time insights.
• Platform Expansion: A need to build a robust initial buyer platform from the ground up to replace legacy processes for both buyers and sellers.
• Third-Party Dependence: Reliance on external rating agencies for viewership data, which the client aimed to replace with an in-house, superior solution.

We assisted the customer in developing their initial buyer platform and executed a comprehensive modernization strategy, including re-engineering the analytics system with real-time data pipelines and implementing AI/ML solutions.

• Marketplace Development: Developed a grounds-up TV advertisement marketplace platform for buyers and sellers, replacing legacy systems.
• Data Engineering: Built data pipelines to process ~30 million records from various channels, developed ETLs to organize raw records, and integrated Tableau for BI visualizations.
• Predictive Analytics (AI/ML): Developed a deep learning and time-series based ensemble model to predict ratings for different demographics and channels, resulting in a new, patented rating engine.
• Feature Development: Built features for buyers to gather budget/ad preferences, developed a billing application based on aired spots, and created business reports for campaign performance analysis.
• QA Automation: Built test automation for the application used by Radio Jockeys.

• Rating engine accuracy: 50% better than third-party solutions in 70% of cases.
• The platform processes approximately 30 million records, enabling immediate business intelligence.
• The platform is widely adopted across the US video advertising ecosystem.
• The AI-driven solution empowers the client to offer superior predictive analytics, strengthening their market position and enabling them to recommend their solution over competitors.

A data management, machine learning, and analytics company sought to transform its existing WXM product into a modern, high-performance “Observability” platform.

The existing platform was a decade old with intricate architecture, suffering from usability and significant performance issues when handling high volumes of data across multiple clusters. The client required a complete re-engineering of the tech stack and user interface to ensure future readiness and enhanced user experience.

• Implemented Environment Hierarchy redesign to address usability and performance issues.
• Developed cost center module for financial governance, to view cluster incurs the cost with cloud resources
• Developed feature to download the chargeback report
• Improved product backend for Private Cloud, Public Cloud & Hybrid Cloud.
• Solved high volume data performance issues across multiple clusters.
• Developed a feature to capture hive metastore information for end users to get more insights about hive engine.

Developed a solution for the complex challenge of displaying driver logs when Spark jobs run in various modes.

• Successfully re-engineered a decade-old platform into a modern Observability product.
• Platform went live for 250 customers within a year.