PRADEEP G

• Conducted in-depth analysis of system issues and implemented effective resolutions to minimize downtime and enhance system reliability.
• Oversaw the architecture, design, and implementation of company-wide platforms and applications, ensuring seamless integration and optimal performance

Product: Moengage Team: Segmentation

• Segmentation V2.0: Played a key role in the development of Moengage's Segmentation V2.0, a critical component of the product. Successfully handled 1.2 Billion monthly active users and processed over 200 billion events from our customers every month.
• UserFlux: Implemented a highly efficient user flux pipeline, enabling smooth indexing of user data into Elasticsearch. This innovative solution now powers user search for a staggering 3 Billion users
• Azure Support: Spearheaded the complete revamp of the segmentation tech stack, shifting to Azure cloud infrastructure. Introduced Trino with Hive-Metastore for efficient querying of ADLS
• Value Suggestion: Led the re-architecture of the value suggestion system, significantly increasing the number of suggestions per attribute to 5k, while simultaneously reducing costs by $2k.
• Optimized YARN resources, leading to substantial cost savings of $20k across data centers.
• Designed UIS system for campaigns to look up user on segment in milliseconds.

Technologies: Kafka, python, java, redis, ES, Athena, Trino, airflow, Scylla, mongo, hive-metastore, glue

• Trained and led team of developers and discussed issues in order to provide resolution and apply best practices
• Managed design efforts and estimates for on schedule delivery of features

Product: Social Marketing, ExInt (Experiences and Intelligence)
Built scalable, concurrent, low latency services for Social Listening tool that ingests billions of posts from social channels.

• Collaborated with PM and other product stake holders to understand pain points of manual labeling, introduced a scalable auto-label service that could label thousands of posts based on predefined rules
• Gathered requirements from several products on handling notifications, developed a unified notification gateway to support different types of notifications
• Designed and developed Twitter Themes to identify topics of conversations using unsupervised learning and spark-NLP

Technologies: Kafka, Ruby, Scala, Spark-NLP, Solr, Scala, Java, SES, Thrift, PostgreSQL, Lucene

Project: Nimesa

Enterprise-class application aware data protection & copy data management solution for applications running on both private and public cloud environment like VMware and AWS

• Been a part of the startup since initial design of the product
• Features Developed: Automatically backup EC2 instances and take application aware snaps on cloud storage systems; verify backups periodically; instant granular recovery of databases, tables, and files from snaps to any EC2 instances

Technologies : RabbitMQ, Java, AWS, SAN