Experience of 14 years on software design and development on Banking Financial systems and Telecom domain .
Experience of 6 years in Big Data technologies . Developed and implemented big data solutions on Databricks and Hadoop platforms using Spark-Scala, pyspark .
Involved in development of data pipelines to process large-scale json datasets using pyspark through databricks notebooks on Azure cloud
Good understanding of databricks concepts such as versioning ,delta table ,time travel . Led the migration from On-prem Hadoop platform to Databricks on azure platform.
Well aware of Agile methodology . Participates in daily scrum meetings , Demo ceremonies.
Worked with container-based technologies like Docker, AKS . Migrated application from VM to Azure Kubernetes Cluster .
Designed and deployed highly available and scalable Kubernetes clusters on AKS. Implemented autoscaling for pods and deployments based on resource utilization.
Hands on Gen-AI technologies including AzureOpenAI, Langchain ,vectordb
Overview
14
14
years of professional experience
1
1
Certification
Work History
Software development Specialist (Lead)
AMDOCS
01.2024 - Current
Application provides summarization and Question & Answer RAG on Tax bills for various states in US
PDF chunking is done using langchain API, embedding is done using AzureOpenAIEmbeddings and stored in milivus vector db
Airflow is used for orchestrating entire application flow
Application deployed on azure Kubernetes cluster.
Senior software engineer
ATTRA INFOTECH
02.2018 - 09.2018
MI issuing is a project to integrate ECS (Electra card services) switch with MasterCard system It handles millions of transactions for ATM, POS, Mobile payments and Internet Daily reports are created using pyspark engine .
Technical Analyst
FISERV
05.2015 - 01.2018
EPOC is a transaction processing software product developed by Fiserv
EPOC host interface application communicates with hosts, processors and networks using various electronic fund transfer message formats including ISO-8583
EPOC handles transactions for ATM, POS, Mobile payments and Internet
Spark-Scala based process is developed to create monthly reports based on daily transactions .
Application Engineer
ATOS WORLDLINE INDIA PRIVATE LIMITED
02.2012 - 04.2015
MAGNUS is EFT (Electronic Fund transfer) acquiring & issuing switch for Point of sale terminals and IPG (Internet Payment Gateway) transactions TCP-IP sockets, UNIX message queues, signaling mechanisms have been used to process multiple transaction requests simultaneously
It is used to acquire, authenticate, route, switch, and authorize financial transactions across multiple channels
Accepts cards of all major associations which include Visa, MasterCard, American express, Rupay.
Cognition solutions pvt ltd
04.2010 - 01.2012
This application is dialog-based application
It's designed to run the various back-office processes
Developed using TCP/IP client - server architecture
The BO Server which is the Back Office Server Application, and it is responsible for all the activities and processing to be done.
Developer, Tech lead
AMDOCS
Guided team for designing and implementing cloud-based data architecture, optimizing performance, and ensuring data integrity
Spark jobs are created in databricks to load daily incremental data in delta tables
Data loading , extraction & transformation is done using Medallion architecture
Implemented REST API to invoke databricks spark jobs
Conducted thorough testing of the new environment to ensure that everything was working as expected and optimized the configuration for performance, cost and reliability.
Developer, Tech lead
AT&T
The purpose of SDP project is to process compensation related data of AT&T dealers & distributors
Payment is generated based on reports provided by SDP
Designed and implemented real time incremental data pipeline to process streaming high volume(~4 million records per hour) json data from 14 different Kafka topics using Apache-Flume and Spark-Scala
Automated ELT (Extraction-loading-transformation) process using Airflow
Moved services to docker -container for high availability (99% uptime).