Senior Cloud & DevOps Engineer working on production systems for banking, capital markets, and healthcare clients under ISO 27001, SOC 2, and HIPAA constraints. Takes engagements from architecture through delivery and operations, usually as sole technical owner. Work spans hybrid migrations, infrastructure as code, microservices, multi-cloud cost optimization ($165K+/yr saved), and applied AI.
Focus
Specializing in high-availability architecture, cost-efficient cloud scaling, and robust CI/CD ecosystems.
Key Impact
"Led infrastructure modularization using Terragrunt — cut onboarding time from 7 days to <1 hour."
"Architected hybrid migrations and cost programs delivering $165K+/year in cloud savings."
"Ran 99.9% uptime Azure SaaS code-review platform and led ISO 27001 & SOC 2 audit readiness."
"Orchestrated org-wide Claude AI adoption (45+/50 devs) and production agentic CI pipelines."
Technical Skills
A comprehensive tech stack optimized for performance, security, and scalability across multi-cloud environments.
Cloud & Infra
Azure
AWS
GCP
AI & LLM
AWS Bedrock
Azure AI Foundry
Multi-Agent Systems
Architecture
Security & Hardening
High Availability
Disaster Recovery
Ops
Cost Optimization
IAC / Automation
Terraform
Terragrunt
Containers
Kubernetes
Docker
Observability
Elastic Stack (ELK)
Prometheus
Grafana
CI/CD
Azure DevOps
AWS Codepipelines
GitHub Actions
Scripting
PowerShell
Bash
Languages
Python
JavaScript
Go
Databases & Data
Postgres
Redis / Valkey
MongoDB
Latest Insights
Sharing my thoughts on Cloud Engineering, DevOps, and the occasional tech deep dive.
SWARM DECENTRALIZED PROTOCOL: Sutradhar (Kernel) connects Tark (Reasoning) and Shilpi (Builder) with Shishya Apprentices. Click any node to trigger Chaos simulation.
Swarm Intelligence & Agents
Setu: Unkillable Autonomous Code
A self-healing coding swarm with functional immortality, fractal memory, and Darwinian token economics. Agents never die, context never resets, and code evolves like a living organism.
DR SIMULATION: Route53 coordinates modular traffic routing. Trigger a regional disaster to observe active failover and replica database promotion in US-WEST-2.
Re-architected per-client/per-environment Terraform setups into a Terragrunt-based modular architecture. Enabled multi-region disaster recovery, simplified infra management, and automated new-client onboarding.
7 days → 1 hour
Onboarding time
Multi-region
DR enabled
TerraformTerragruntAWSCI/CDDisaster Recovery
Private RepoCode is private for this institutional/corporate project
Architected Azure infra (App Services, Postgres, OpenAI integration) for an AI code-analysis product to achieve high availability and simplified deployments using Dockerized workflows.
99.9%
Uptime
↓90%
Deployment errors
AzureDockerPostgresOpenAIHigh Availability
Private RepoCode is private for this institutional/corporate project
ELASTIC_STACK_METRICS
INGEST_RATE:14.8M/week
ACTIVE_NODES:3 ONLINE
ELASTALERT_STATUS
12% CPU
NO_THREATS_DETECTED
🔗 LOG_INGESTION_PIPELINE
📊 KIBANA METRICS HUD
CLUSTER_HEALTH
GREEN_OPERATIONAL
JVM_HEAP_USAGE12.4%
INDEXED_DOCS_COUNT
847,203,194+1,240 docs/s
[INFO] Elastic system starting up...
[INFO] Elasticsearch cluster state: GREEN (3/3 nodes operational)
[INFO] Ingestion index mapping compiled successfully.
KIBANA OBSERVABILITY: Continuous logs stream into Elasticsearch indexers. Trigger an anomaly to flood index loads and watch JVM heap spikes.
Logging & Monitoring at Scale
Elastic Stack Observability Platform
Deployed Elastic Stack from scratch and designed ingestion pipelines to collect ~15M logs/week. Built dashboards and alerts (ElastAlert2) to improve monitoring and incident response.
15M
Logs ingested/week
3 core system dashboards
Dashboards
Elastic StackElastAlert2KibanaSRE
Private RepoCode is private for this institutional/corporate project
PLACEMENT_ALGO_DRIVE
DRIVE_STATE:IDLE
AVG_OFFER:0.0 LPA
OFFERS_MADE:0 secured
PLACEMENT_RATE
0%
NIT_CDC_ANALYTICS
NIT MATCHMAKER: Floating nodes represent students (left) matching with eligibility partners (right). Start drive to observe rapid placement algorithms matched.
Education Portal & Analytics
Training & Placement Cell Portal (NIT Surat)
Built a student placement portal that automated application workflows and provided data insights for 800+ students; reduced process time by ~80%.
800+
Users impacted
↓80%
Process time
Node.jsMySQLJavaScript
Private RepoCode is private for this institutional/corporate project
Architected and delivered a hybrid Azure migration for a global bank’s corporate actions platform, covering secure on-premises connectivity, infrastructure consolidation, and licensing optimization, reducing client operating costs by $107K/year.
Pitched and led an AI-assisted Azure CIS security audit across eight domains for a post-incident financial services client, remediating every finding, including a stale privileged access path left over from earlier project work.
Cloud & DevOps Engineer
June 2024 - June 2026
Fintech Global Center (GIFT City, India)
Cut new client onboarding from 7 days to under an hour by leading a Terragrunt migration that turned branch-based Terraform into a modular, multi-region, DR-tested architecture.
Re-architected a 25-year-old VB6 monolith into five .NET microservices for a US regional bank, owning the architecture, CI/CD pipelines, and a shared Angular component library adopted company-wide; the modernization cut response times 60%.
Pitched and built a multi-tenant client operations control plane for cross-account AWS secrets management, with drift detection, audit trails, and feature-flag governance. Delivered end to end solo and now running in production.
Led AI adoption across a 50-person organization: migrated the company to the Claude Team plan under a zero-data-retention agreement, drove weekly active usage to 45+ of 50, built custom plugins and skills, and rolled out agentic CI workflows that open guideline-compliant pull requests behind approval gates.
Architected a HIPAA-aligned patient intake system on AWS using ECS, RDS, and Bedrock. Rejected the original diagnosis-driven scope and redirected it toward dense structured summaries that support physician judgment rather than replace it.
Ran the Azure platform for a public SaaS code-review product (App Service, PostgreSQL, Azure OpenAI) at 99.9% uptime for 22 months, eliminated non-deterministic deployments by containerizing the release path, and owned cloud-domain evidence and remediation for ISO 27001:2022 and SOC 2 Type II audits.
Ran a multi-cloud cost optimization program across AWS and Azure covering Graviton migration, EKS capacity tuning, unused database cluster removal, a Redis to Valkey migration, and a runaway Lambda condition surfaced through cost analysis, cutting spend by $58K/year, about 35% of total estate spend.
Cloud & DevOps Intern
Jan 2024 - May 2024
Fintech Global Center (GIFT City, India)
Deployed a self-managed Elastic Stack from scratch, a 3-node cluster with ILM and custom ingest pipelines handling up to 15M logs weekly, integrated ElastAlert 2 alerting that caught a silent ILM failure before it filled the cluster, and built 3 monitoring dashboards.
Introduced a branching and release-tagging strategy where none existed, then built Azure CI/CD pipelines on top of it, including a hotfix workflow that cut emergency releases from 3-4 hours of manual work to under 30 minutes.
Software Development Intern
May 2023 - July 2023
John Deere (Pune, India)
Migrated a 5M-row database from German to English during a MySQL to AWS RDS transition using enum-based deterministic mapping; reduced upload errors by adding type and range constraints at the data-entry layer.