Jul 2021 – Sep 2024 · 3 yrs 2 mos · Bengaluru, Karnataka, India

• Key Projects:
• Microservices Event Generation and Storage in BigQuery
• Engineered an event framework for developers to generate events, catering to audit and business data, and ML model requirements.
• Facilitated data transmission from JAVA code to Kafka from diverse microservices after validation with schema of BigQuery table.
• Streamed data from Kafka to GCS using Vector.dev and used BigQuery load jobs API in JAVA to load data from GCS to BigQuery.
• Elevated developer experience by crafting a React.JS UI, offering a platform for creating, editing, and visualizing load jobs in BigQuery.
• CIBIL Data Integration
• Orchestrated CIBIL API integration, extracting customer scores, personal details, and loan history in JSON format.
• Engineered a resilient microservice in Play Framework, utilizing WS library for API calls with a retry mechanism.
• Leveraged Jackson library with custom deserializers to parse JSON data into Java classes, enabling storage in MySQL DB.
• Clickhouse Integration and Data Sharding
• Utilized Kafka for real-time data capture from microservices for ML feature calculation.
• Evaluated Clickhouse data insertion methods using Vector.dev, Clickhouse Kafka Engine, and Clickhouse JAVA API.
• Implemented sharding for data distribution across Clickhouse nodes, optimizing scalability and using Kafka Engine for insertion.
• Task Scheduler
• Engineered a JAVA scheduler, enabling developers to execute any method within any Java class at defined intervals using cron expressions.
• Integrated scheduler in Play Framework as a module and used ExecutorService API to check methods to be executed every minute.
• Leveraged Kafka for method execution requests from diverse microservices and used reflection and multi-threading to execute methods.

May 2019 – Jul 2019 · 2 mos · Hyderabad, Telangana, India

• Objective:
• To analyse reliability data to obtain its predictability and improve operations and maintenance of airplanes
• Approach:
• Analysed various types of two years reliability data by making data frames using Pandas library in Python
• Implemented Natural Language Processing using Nltk library to analyse discrepancy text using Tf-idf score
• Predicted cancellations and time of delays of an airplane by discrepancy text using 4 machine learning algorithms
• Results:
• Achieved an accuracy score of 95% using Multinomial Naive Bayes algorithm in predicting time of delays
• Displayed major reasons of discrepancies using matplotlib and purposed analysis of maintenance action for them