NITHEESH CHANDRA SHAMANTHULA

Backend & Systems Engineering Contributions

• Started Working on one of our inhouse Python FastAPI based tool which is one stop tool for SREs. Written in ASYNC Manner for IO and Network Calls. It uses LLMs, Vector Embeddings & Databases, Redis and PostgreSQL.

Backend & Systems Engineering Contributions

• Built Python-based automation for Kafka topic management, enabling bulk operations via optimized CSV ingestion. The script creates/updates 1000s of topics in one single API call to Kafka. All topics to be created/modified are source controlled.
• Written a Simple Golang Service for generating Cloudfront Signed URLs for various s3 buckets and paths.

Infrastructure & DevOps Contributions

• Have set up a complex networking setup, which included adding a firewall between private subnets, site-to-site connections, and VPC.
• Have set up a new vertical infrastructure from scratch, including the OCI new tenancy setup, networking, and K8s cluster setup.

Backend & Systems Engineering Contributions

• Designed and built a Golang-based DMARC authentication service, processing 10M+ emails/day, with strong adherence to RFC standards
• Designed and Built a Python Probe service for SLA monitoring (latency, uptime, functional checks). Previously we used to use Site24x7, but it was not reliable when it comes to functionality and latency checks. So this service was written in such a way that it can be hosted anywhere and it can generate the SLA metrics for any partner on top of our systems. Most parts were written as a module for reusability across multiple projects where there is need to doing health/functionality checks in the their services. Pytest was used for writing test cases.
• Python IP blocking Service using Redis + Elasticsearch for real-time rate limiting of incoming and outgoing mails
• Created NodeGroup upgrade automation scripts (Python) reducing upgrade time from days to hours. We have most of the dev services written in JAVA and run on shared instances, and when we used to do rolling upgrades of nodes one by one, we used to see multiple services restart due to Higher CPU usage at the JVM Startup. This Helper Tools help by rollout restarting each service one by one. Later on have added support for rollout restart of Elasticsearch Clusters since its not straight forward and also custom rollout restart mechanism for couple of other services which were under Classic Load Balancer.
• Developed Python Helper Scripts for Effective Log Parsing parallely using multiprocessing
• Profiled Couple of Python/Golang services using pyroscope in general, and memray for python memory profiling.
• Extended and customized open-source exporter (Postfix Exporter) for internal observability needs

Infrastructure & DevOps Contributions

• Managed large-scale Kubernetes clusters (100+ nodes, 100+ services), including migrations and upgrades
• Designed observability pipelines (metrics, logs, traces) across mail infrastructure
• Performed database upgrades and maintenance (MySQL, Elasticsearch)
• Contributed to Opensource Envoy Gateway (v1.2.0) via Helm chart improvements