As a Software Engineer in the Security organization at Salesforce, I develop and deploy machine learning systems for real-time threat detection. I specialize in designing distributed systems that support real-time model training and inference at scale. My work includes building ML pipelines for detecting anomalous and malicious behavior, deploying models into production environments, and developing services that enable enterprise-scale data governance, including automated data quality validation and data lineage tracking across large datasets.
Posts
* Software Engineer on the Security team at Salesforce, building and operating ML-powered solutions used for Salesforce-wide security.
* Currently building a real-time ML platform for near-real-time threat detection (<60 seconds), supporting production use cases such as login anomaly and data exfiltration detection.
* Have worked across the ML platform lifecycle, including data quality, data lineage, pipeline reliability, and data availability for production ML systems.
* Strong experience with distributed systems, scalable data/ML pipelines, and platform engineering, with a focus on building reliable and production-ready systems.
* Experienced in working on large-scale engineering problems and collaborating across teams to take systems from development to production.
Experience
Software Engineer (MTS) at Salesforce Jun 2024 to now - Bangalore Automated end-to-end remediation of stolen credentials and leaked API secrets across Salesforce Core and Marketing Cloud by building a fault-tolerant service that consumes threat detections from Kafka and orchestrates credential verification followed by password reset, credential suspension, or IP blocking through authenticated cross-cloud APIs. Transformed a batch-based ML threat-detection system into a real-time streaming platform, reducing detection latency by 96% (30 min → <60 sec) and event loss by 90%, by re-architecting centralized Spark batch jobs into a distributed, fault-tolerant pipeline—building the per-region log-consumer and Kafka-based egress service
SWE Intern at Salesforce May 2023 to Jul 2023 - Bangalore Developed a proof of concept for integrating Feast as a feature store for streaming detection jobs, enabling low-latency storage and retrieval of stateful features for real-time model inference.
Education
Indian Institute of Technology (IIT), Hyderabad B.Tech, Artificial Intelligence 2020 to 2024 Relevant Coursework: Data Structures and Algorithms, DBMS, Discrete Maths, Operating Systems, Computer Networks, Foundations of ML, Deep Learning, Image and Video Processing, NLP, Reinforcement Learning, Explainable AI