Arnab Gupta

Projects: Central Cache Service, Prioritization Framework

Tech Stack: Java, C, Redis, Spring Boot, MySQL, GitLab, Kubernetes

• Design and development : Actively involved in the design, development, deployment, testing automation for the central cache system and prioritization engine.
• Unit Testing, Load Testing and Integration Testing Automation : Collaborated with project managers, developers, quality assurance, and clients to promptly address and resolve technical issues. Wrote a custom test framework for testing prioritization framework test cases to improve coverage > 80%.
• Central Cache Service: Designed a central cache API which supports upwards 50000 rps. This allowed various verticals to utilize caching in a systematic way. Max recorded usage for this micro-service for a day was around 34.5 crore hits. Also helped migrate the cache service to kubernetes cluster from ec2 instances.
• Prioritization Framework : Built a system from scratch that allows our consumers to provide guaranteed rps to our end customers without hampering the overall efficiency of the system and choking other customers. The engine is a custom redis module which is build on top of Redis Module APIs in C. Overall the system recorded was upwards of 28000 rps.
• CI/CD Pipeline Revamp : Revamped the existing CI/CD pipeline, to make it reusable for multiple projects.

Projects: Fraud Management System, Revenue Assurance, Financial Reconciliation

Tech Stack: Java, Postgres, Spring Boot, Redis, MongoDB, RabbitMQ, Spark, Hadoop

• Design and development : Actively participated in the end-to-end design, development, and deployment of three distinct projects, each consisting of multiple microservices. Focused on robust backend development, built them from scratch, and integrated support for asynchronous HTTP calls from the User Interface.
• Unit Testing and Integration Testing : Collaborated with project managers, developers, quality assurance, and clients to promptly address and resolve technical issues. Wrote unit test cases to improve coverage > 70%
• Facial Comparison Tool : Pioneered the creation of this tool as part of the Fraud Management System, utilizing ML models to compare image similarities. Aided in the semi-automation of the KYC system in Nigeria and Rwanda through a generic API interface. The API was successfully scaled to handling 5lakh requests per hour.
• Financial Reconciliation Tool : Designed and built from the ground up, allowing business users to detect payment-related anomalies. Achieved an accuracy rate of 80% to 110% for various reconciliation datasets during its beta launch.
• Revenue Assurance Enhancements : Enhanced roll-up cubes ETLs using Spark to optimize metadata generation and cube selection, contributing to significant improvements in the existing pipeline. Also built KPI engine to automate report generation on various datasets.
• CI/CD Pipeline Revamp : Revamped the existing CI/CD pipeline, established development and production environments, significantly improving efficiency and deployment processes.

Projects: Revenue Assurance

Tech Stack: Java, Postgres, Spring Boot, Redis, MongoDB, RabbitMQ

• Helped build scalable APIs and API aggregators from scratch using spring boot framework for the big data team for data visualization and aggregation dashboards.
• Designed, conceptualized and contributed to the development of a monolithic rule engine and extending the service to micro-service architecture.
• Helped in ensuring smooth deployment of APIs across multiple OPCOs.
• Participated in code reviews and black box testing of the developed micro services.
• Created unit testing and created overall testing plans for the APIs and other microservices.
• Enhanced and optimized query performances and configuration load times with collaboration of platform team and decreased API response time by 400%.

Projects: Search API

Tech Stack: Groovy, Kafka, Mongo DB, Elastic Search.

• Enhanced existing APIs to expose new features to user dashboard.
• Contributed to feature enhancement and stability.
• Created Loader classes to move configs to mongo db.
• Added new dashboard to existing configuration management tool to configure new jobs

Projects: Notification Engine, Alerts Engine, Reporting Tool

Tech Stack- MySQL, Elasticsearch, Redis, Google Big Query, RabbitMQ. Java 8

• Created browser campaign support for the internal campaign tool to schedule campaign for a specified user base.
• Integrated changes in the campaign tool for creating segments using predefined options that dynamically generates query based on previous user activity data stored in Google BigQuery.
• Created changes both on internal campaign tool and notification system for storing, updating and displaying campaign run metrics on campaign tool web interface.
• Tested troubleshooting methods, devised innovative solutions, and documented resolutions for inclusion in knowledge base for support team use.
• Made changes to existing SMS campaign feature in campaign tool to support exclusion of blacklisted and invalid users as part of cost optimization.
• Enhanced the existing alert service by using the existing unsubscribe data to target the valid user base.

Projects: Reporting Tool

Tech Stack- MySQL, Google Big Query, RabbitMQ. Java 8

• ETL workflow orchestration using airflow to generate CSV file format using BigQuery data and upload to external vendor sftp location.
• Optimized data fetch from MySQL databases by reducing usage of system memory by process.
• Created an ETL pipeline to support data transfer from Big Query to Amazon S3 bucket.
• Provided support to reporting system by closely monitoring the system and resource critical alerts.
• Wrote ansible automation scripts for automating the process of deployment.