Subho D. profile photo 1 on Tal

Subho D.

Senior Platform & DevOps Engineer @ OpenGov | SRE · Cloud Infrastructure · MLOps/AIOps · FDE| AWS · Kubernetes(EKS) · Terraform · GitOps (Argo CD) | Scaling 30+ EKS clusters & GenAI platform infra | Freelancer.

Pune Division

9+ years of work experience

OpenGov Inc.
Feb 20242 yrs 6 mosPresent

• Partnered directly with engineering leadership and product teams to bridge the gap between business requirements and technical implementation and translated complex needs into architectural POCs, including a gVisor (runsc) sandbox runtime on EKS for secure AI-agent workloads and a Snowflake SPCS evaluation for in-database AI gravity. •Standardized organization-wide CI/CD pipelines following a pure GitOps model; utilized GitHub Actions for CI (developing reusable workflow templates), Argo CD for CD, and Kargo for automated PR-driven multi-stage promotion across 163+ warehouses and 58 projects, eliminating 5-minute polling latency. • Delivered the production cutover of the OpenGov AI Agents platform — AWS account wiring, a dedicated EKS node group, an Amazon Bedrock Knowledge Base RAG ingestion pipeline in Terraform, S3-to-VPC-private Lambda integration, CloudWatch subscription-filter observability, OpenTelemetry/Grafana dashboards, and Opsgenie on-call setup. • Own Kargo as a fleet-wide GitOps promotion platform across 163+ warehouses and 58 projects; upgraded through three major versions (v1.5 to v1.10) with zero downtime and eliminated 5-minute polling latency for 13 teams via a GitHub webhook + EventBridge receiver pattern. • Own the Amazon EKS cluster lifecycle, including an org-wide Kubernetes 1.32-to-1.33 upgrade (control plane, managed node groups, kube-proxy, VPC CNI, CoreDNS, EBS CSI driver) and a greenfield production Kubernetes 1.34 cluster built end-to-end across three Terraform repositories. • Reduced EKS spend by ~$100K/year through Robusta KRR-driven pod right-sizing, and cut cluster compute further ~30% by pairing KEDA event-driven autoscaling with right-sized managed node groups. • Reduced Terraform Cloud plan time 5x (38 to 8 minutes) by tuning provider rate limits and parallelism.
OpenGov Inc.
Aug 20221 yr 7 mosFeb 2024

• Designed a PR-driven image-promotion pipeline using GitHub Actions, Argo CD, and Kustomize for integration, staging, and production releases across application teams; eliminated manual release toil while preserving review/approval gates with one-click reversible rollbacks. • Led the ECS-to-EKS migration of a major business-suite backend — translating ECS task definitions into Kustomize-managed Deployments and wiring IRSA, ExternalSecrets-backed Aurora/Redis credentials, ingress, HPA, and Argo CD ApplicationSet entries; reused the pattern for the n8n automation platform across three environments with zero downtime. • Built the organization's unified database-access plane on Teleport across 140+ Aurora clusters and 215+ RDS instances, replacing static credentials with IAM-auth DB roles and Okta group-to-role mapping, with full session- audit coverage.
Coditas
Feb 20202 yrs 7 mosAug 2022

pune

•Designed and implemented secure three-layer VPC architectures (Public, Private, Data) and established scalable networking foundations on AWS. • Led zero-downtime cloud migrations, re-platforming 50+ services from EC2 to Kubernetes for global EdTech and HealthTech clients; designed secure three-layer VPC architectures and CI/CD pipelines with Jenkins and GitLab CI. • Reduced MTTR by ~40% by building a centralized observability stack with Prometheus, Grafana, and Sensu; enabled blue-green and canary deployments by containerizing legacy monoliths to Amazon ECR. • Migrated CI/CD from Bitbucket Pipelines to self-hosted Jenkins with a master/agent architecture on AWS, reducing CI costs while mentoring teams on Infrastructure as Code adoption.
Aurus Inc USA and Aurus Tech Pvt Ltd, Pune India
Aug 20172 yrs 7 mosFeb 2020

pune

• Reduced Docker build times by ~60% by standardizing development environments and Jenkins-based CI; hardened MariaDB, Apache, and SFTP on Linux to reduce security incidents. • Automated routine system administration and log-management workflows with Bash, saving engineering hours and improving proactive alerting.
GitHub
@subhodas09

Subho’s AI Stack

antigravity
antigravity
claude code
claude code
claude
claude
codex
codex
cursor
cursor
openai
openai

Education

Yeshwantrao Chavan College of Engineering - YCCE

Bachelor of Engineering (BE), Electrical, Electronics and Communications Engineering

Yeshwantrao Chavan College of Engineering - YCCE

20134 yrs2017
kendriya vidyalaya

kendriya vidyalaya

200112 yrs2013

Join Subho and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Subho D. on Tal

Headline

Senior Platform & DevOps Engineer @ OpenGov | SRE · Cloud Infrastructure · MLOps/AIOps · FDE| AWS · Kubernetes(EKS) · Terraform · GitOps (Argo CD) | Scaling 30+ EKS clusters & GenAI platform infra | Freelancer.

About

Senior Platform & DevOps Engineer with 8+ years across Cloud Infrastructure, Site Reliability Engineering (SRE), and DevOps. I currently run a ~$1.4M/year AWS production estate for a govtech multi-tenant SaaS serving U.S. state and local governments (SOC 2 / StateRAMP). What I do — I operate 30+ Amazon EKS clusters, 350+ Aurora/RDS databases, and 100+ services at 99.9% availability. I own GitOps at scale with Argo CD and Kargo (163+ warehouses, 58 projects), Infrastructure as Code with Terraform, and the end-to-end Kubernetes cluster lifecycle — from org-wide version upgrades to greenfield production builds. Recent impact — • Delivered production-grade GenAI infrastructure: an Amazon Bedrock Knowledge Base RAG pipeline and a gVisor-isolated sandbox runtime on EKS for safely running AI-agent workloads. • Cut ~$100K/year in EKS spend via KRR-driven right-sizing and KEDA event-driven autoscaling. • Improved Terraform plan time 5x (38 to 8 minutes) and removed manual release toil with PR-driven, one-click- reversible deployment pipelines. • Built a unified, fully audited database-access plane on Teleport across 350+ databases with Okta-gated IAM authentication. Focus areas — Platform Engineering, DevOps, Site Reliability Engineering (SRE), Cloud Infrastructure, MLOps / AIOps, and GenAI platform infrastructure. Toolkit — AWS (EKS, Aurora/RDS, Lambda, Bedrock, SageMaker, EventBridge, IAM Identity Center, VPC), Kubernetes, Terraform & Terraform Cloud, Argo CD, Kargo, Helm, Kustomize, GitHub Actions, KEDA, Prometheus, Grafana, OpenTelemetry, CloudWatch, Opsgenie, Teleport, and Okta. Currently pursuing AWS Certified Solutions Architect – Professional and Certified Kubernetes Administrator (CKA). Open to Senior Platform, DevOps, SRE, Cloud Infrastructure, and MLOps/AIOps roles — remote, hybrid, or on-site. Feel free to connect.

Location

Pune Division

AI stack

antigravity

claude code

claude

codex

cursor

openai

Work experience

Senior DevOps Engineer at OpenGov Inc., Feb 2024 - Present · 2 yrs 6 mos, • Partnered directly with engineering leadership and product teams to bridge the gap between business requirements and technical implementation and translated complex needs into architectural POCs, including a gVisor (runsc) sandbox runtime on EKS for secure AI-agent workloads and a Snowflake SPCS evaluation for in-database AI gravity. •Standardized organization-wide CI/CD pipelines following a pure GitOps model; utilized GitHub Actions for CI (developing reusable workflow templates), Argo CD for CD, and Kargo for automated PR-driven multi-stage promotion across 163+ warehouses and 58 projects, eliminating 5-minute polling latency. • Delivered the production cutover of the OpenGov AI Agents platform — AWS account wiring, a dedicated EKS node group, an Amazon Bedrock Knowledge Base RAG ingestion pipeline in Terraform, S3-to-VPC-private Lambda integration, CloudWatch subscription-filter observability, OpenTelemetry/Grafana dashboards, and Opsgenie on-call setup. • Own Kargo as a fleet-wide GitOps promotion platform across 163+ warehouses and 58 projects; upgraded through three major versions (v1.5 to v1.10) with zero downtime and eliminated 5-minute polling latency for 13 teams via a GitHub webhook + EventBridge receiver pattern. • Own the Amazon EKS cluster lifecycle, including an org-wide Kubernetes 1.32-to-1.33 upgrade (control plane, managed node groups, kube-proxy, VPC CNI, CoreDNS, EBS CSI driver) and a greenfield production Kubernetes 1.34 cluster built end-to-end across three Terraform repositories. • Reduced EKS spend by ~$100K/year through Robusta KRR-driven pod right-sizing, and cut cluster compute further ~30% by pairing KEDA event-driven autoscaling with right-sized managed node groups. • Reduced Terraform Cloud plan time 5x (38 to 8 minutes) by tuning provider rate limits and parallelism.

DevOps Engineer at OpenGov Inc., Aug 2022 - Feb 2024 · 1 yr 7 mos, • Designed a PR-driven image-promotion pipeline using GitHub Actions, Argo CD, and Kustomize for integration, staging, and production releases across application teams; eliminated manual release toil while preserving review/approval gates with one-click reversible rollbacks. • Led the ECS-to-EKS migration of a major business-suite backend — translating ECS task definitions into Kustomize-managed Deployments and wiring IRSA, ExternalSecrets-backed Aurora/Redis credentials, ingress, HPA, and Argo CD ApplicationSet entries; reused the pattern for the n8n automation platform across three environments with zero downtime. • Built the organization's unified database-access plane on Teleport across 140+ Aurora clusters and 215+ RDS instances, replacing static credentials with IAM-auth DB roles and Okta group-to-role mapping, with full session- audit coverage.

DevOps Engineer at Coditas, Feb 2020 - Aug 2022 · 2 yrs 7 mos, pune, •Designed and implemented secure three-layer VPC architectures (Public, Private, Data) and established scalable networking foundations on AWS. • Led zero-downtime cloud migrations, re-platforming 50+ services from EC2 to Kubernetes for global EdTech and HealthTech clients; designed secure three-layer VPC architectures and CI/CD pipelines with Jenkins and GitLab CI. • Reduced MTTR by ~40% by building a centralized observability stack with Prometheus, Grafana, and Sensu; enabled blue-green and canary deployments by containerizing legacy monoliths to Amazon ECR. • Migrated CI/CD from Bitbucket Pipelines to self-hosted Jenkins with a master/agent architecture on AWS, reducing CI costs while mentoring teams on Infrastructure as Code adoption.

Technical Support Engineer at Aurus Inc USA and Aurus Tech Pvt Ltd, Pune India, Aug 2017 - Feb 2020 · 2 yrs 7 mos, pune, • Reduced Docker build times by ~60% by standardizing development environments and Jenkins-based CI; hardened MariaDB, Apache, and SFTP on Linux to reduce security incidents. • Automated routine system administration and log-management workflows with Bash, saving engineering hours and improving proactive alerting.

Education

Bachelor of Engineering (BE), Electrical, Electronics and Communications Engineering, Yeshwantrao Chavan College of Engineering - YCCE

kendriya vidyalaya

Public links

github: https://github.com/subhodas09