Headline

Goldman Sachs | IIT Kharagpur

About

Backend Engineer at Goldman Sachs | 4+ years building real-time data pipelines and distributed systems for compliance platforms in financial services. I design and ship high-throughput streaming infrastructure using Apache Flink, Kafka, Elasticsearch, and MongoDB — processing millions of events daily with sub-second latency. My work powers trading restriction enforcement, trade surveillance, and regulatory reporting across the firm. What I do: real-time stream processing • stateful event-driven architectures • self-healing data reconciliation • batch-to-streaming migrations • performance optimization at scale. Previously built trade surveillance detection models using Hadoop MapReduce and HBase, improving detection accuracy by 30% while reducing false positives. I care about reliability, clean abstractions, and building systems that scale. Most of my work lives at the intersection of streaming data, distributed systems, and financial compliance. Tech: Java · Apache Flink · Kafka · Elasticsearch · MongoDB · Hadoop · HBase · DB2 · Docker · Kubernetes

Location

Mumbai Metropolitan Region

AI stack

claude code

claude

gemini

github copilot

mcp

openai

Work experience

Associate at Goldman Sachs, Jun 2022 - Present · 4 yrs 3 mos, Bengaluru, Karnataka, India, * Built and deployed 13 real-time Apache Flink pipelines, replacing batch ETL with sub-second data propagation across MongoDB, Elasticsearch, and Sybase. * Engineered scalable data enrichment and reconciliation frameworks processing 600M+ records, reducing data issue resolution from days to hours through automated validation and self-healing pipelines. * Diagnosed and resolved a production-critical concurrency issue, restoring full cluster stability and improving reliability of distributed data processing. * Led infrastructure modernization initiatives including SFTP migration, RHEL 7→8 upgrade, static code analysis remediation, and vault-based secret management. * Developed trade surveillance models for equities, fixed income, and FX across global business units, improving anomaly detection accuracy by 30% while reducing false positives. * Migrated legacy Java batch applications to Hadoop MapReduce, achieving ~10× faster processing for millions of daily trade records.

Software Engineer at AppViewX, May 2021 - Aug 2021 · 4 mos, • Created Clusters using Kubernetes and worked on creating multiple pods, replication controllers, replica sets by writing Yaml files • Automated the process of plugin upgrade, reduced the installation time by avoiding repetition and increased the performance by 50% • Designed a code to dynamically create password for UI kibana dashboard and display when install script is run on the virtual machine • Automated the deployment of 30+ cloud servers and reduced costs by ~$3,000 each month by eliminating the unnecessary servers

Deep Learning Intern at venturit., Jul 2020 - Sep 2020 · 3 mos, India, • Designed a multi-class deep learning model capable of identifying Breast Cancer with accuracy up to 91% • Built the model by transfer learning using award-winning architectures for computer vision like ResNet-50, VGG16, Inception

General Secretary at Nehru Hall of Residence, IIT Kharagpur, May 2020 - Aug 2021 · 1 yr 4 mos, Kharagpur, West Bengal, India

Machine Learning Intern at The Rolling Circle, May 2020 - Jul 2020 · 3 mos, Mumbai, Maharashtra, India, • Part of the team to develop an MVP(Minimum Viable Product) for creating human-like voiceover using Speech Synthesis • The project was developed using Google Tacotron as base and NLP(Natural Language Processing) and Deep Learning

Education

BTech - Bachelor of Technology, Indian Institute of Technology, Kharagpur

Secondary School Certificate, St. Lawrence High School Nashik