This site is a work in progress — some sections are incomplete.
$

git log --oneline --graph

Timeline of milestones

8a3f91 feat (platform): scaling a multi-tenant conversational-AI platform to 15k CCU Current
  • Took CCU ceiling from 800 → 13k+ (simple bots) and 4.5k (moderate bots)
  • Eliminated bottlenecks tier-by-tier: RabbitMQ → Redis → MongoDB → NFS
  • Cut cost-per-scaling-event ~27% via workload-specific node groups + KEDA
7c2b14 perf (rabbitmq): eliminated RabbitMQ as the primary scaling bottleneck Pillar P04
  • ha-all → ha-two policy; cluster load avg 150+ → ~15
  • Error rate 6% → 0.05% after queue-class rebalancing
  • Re-topologised 1×32 → 4×8 clusters for blast-radius isolation
6d1a92 chore (autoscaling): replaced CPU HPA with KEDA + per-pool saturation metrics Pillar P05
  • CPU HPA was wrong default for a Node.js event-loop workload
  • Designed workload-specific node groups (app / ml / rmq / mongo)
  • Asymmetric cooldowns kept cost growth sub-linear with CCU
5e4c83 feat (k8s): led zero-downtime VM → Kubernetes migration across AWS + Azure Pillar P02
  • Stateful conversational-AI platform off VMs onto EKS + AKS
  • Eliminated cascading restarts via probe redesign (startup / readiness / liveness)
  • Produced golden-path manifests adopted across multiple teams
4f1d77 feat (data): built the analytics off-ramp: Mongo → Kafka → Glue → Hudi Pillar P03
  • Athena/Trino query layer over Hudi tables on S3
  • Resolved recurring Glue failure modes (OOMs, schema drift, small-files)
  • Caught a silent sync-stop bug that was eroding data freshness
3a8e60 init (career): started @ Kore.AI · Software Engineer, Platform Jun 2023
$

cat ~/.config/stack.yaml

Things I use day-to-day, and things I'm exploring.

Languages

JavaScriptTypeScriptPythonC++Bash

Runtimes & Containers

Node.jsDocker

Orchestration

KubernetesHelmIstioKEDA

Cloud

AWS (EKS, EC2, ECR, S3, ElastiCache)Azure (AKS)Terraform

Data & Messaging

MongoDBRedisRabbitMQKafka

Analytics & Lakehouse

Apache HudiAWS GlueTrino / AthenaClickHouse

Observability

PrometheusGrafanaLokiGrafana AlloyOpenTelemetry

Performance

JMeterk6

CI / CD

HarnessTerraform CloudArtifactoryGitHub Actions

Exploring

GoRustPostgresGCPeBPF
$

cat experience.md

7 roles · industry + research

Software Engineer · Platform (Infrastructure) Performance

Kore.AI · Full-time

Jun 2023 – Present · ~3 yrs · Hyderabad, Telangana, India

  • Owning performance engineering for a multi-cloud platform serving 15k+ concurrent users across 100+ services.
  • Drove infrastructure fine-tuning, Kubernetes platform work, and storage solutioning across AWS and Azure.
  • Built incident-response playbooks and observability-driven automation that cut MTTR meaningfully.
KubernetesSystem TestingPerformanceStorage

Graduate Research Assistant

Computer Systems Group · Full-time

May 2020 – Jul 2024 · 4 yrs 3 mos · Hyderabad, Telangana, India

  • Researched under Dr. Deepak Gangadharan in the Computer Systems Group, IIIT Hyderabad.
  • Published thesis: "Data Age Formulation and Analysis in Real-Time Embedded Systems — Fault-Tolerant and Thermal-Aware Perspectives".
Real-Time SystemsEmbedded SystemsResearch

Teaching Assistant (3 terms)

IIIT Hyderabad · Part-time

Jan 2021 – Apr 2023 · ~13 mos total · Hyderabad, Telangana, India

  • Computer Architecture & Design — Spring 2023 — under Prof. Praveen Paruchuri.
  • Human-Computer Interactions — Fall 2022 — under Prof. Nimmi Rangaswamy.
  • Computer System Organisation — Spring 2021 — under Dr. Deepak Gangadharan.
TeachingComputer ArchitectureHCI

Research Assistant

iRaste, INAI · Part-time

Sep 2022 – Mar 2023 · 7 mos · Hyderabad, Telangana, India

  • Collaborated with Intel India to implement ADAS Level-2 systems for public-transport buses across Maharashtra and Telangana.
  • Explored the HCI angle — driver willingness and adaptability — under Prof. Nimmi Rangaswamy.
  • Initiative funded by the Ministry of Transportation, India; aimed at road safety and operational efficiency.
ADASHCIResearch

Student Teacher

iHub-Data · Part-time

Jun 2022 – Mar 2023 · 10 mos · Hyderabad, Telangana, India

  • Designed and delivered an AI/ML course for working professionals, DRDO freshers, and management personnel.
  • Built teaching materials, hands-on assignments, and lecture content tailored to a non-academic audience.
TeachingAI / ML

Research Intern

Samsung R&D Institute India — Bangalore · Internship

May 2022 – Aug 2022 · 4 mos · Bengaluru, Karnataka, India

  • Worked on Samsung Neural Accelerator Platform (SNAP) to improve ML model performance on the chipset.
  • Proposed a pre-processing approach that improved performance by 0.3%.
OpenCLC++ML Acceleration

Software Engineer Intern

Indriyn Data Analytics pvt ltd · Internship

Feb 2020 – Apr 2020 · 3 mos · Hyderabad, Telangana, India · Hybrid

  • Built energy-consumption prediction using Facebook Prophet with caching for accurate forecasts.
  • Integrated the model into a user-friendly MERN-stack web application.
ProphetMERN StackForecasting
$

cat education.md

M.S. by Research, Real-Time Embedded Systems

2020 – 2024

IIIT Hyderabad

Thesis: "Data Age Formulation and Analysis in Real-Time Embedded Systems — Fault-Tolerant and Thermal-Aware Perspectives".

B.Tech (Hons), Computer Science & Engineering

2018 – 2022

IIIT Hyderabad

Dean's List · Research Award · Teaching Assistant

$

ls publications/

2 entries · academic output

Data Age Formulation and Analysis in Real-Time Embedded Systems — Fault-Tolerant and Thermal-Aware Perspectives

MS by Research Thesis

IIIT Hyderabad Jul 6, 2024

Checkpointing-Aware End-to-End Data Age Analysis of Task Chains under Transient Faults

Conference Paper

IEEE Jun 12, 2024

Proposes an analytical framework for calculating data age in real-time task chains under transient faults; quantifies how checkpointing interference affects predictability.