Loading repository data…
viveksinghfx / repository
Devop_Portfolio
DevOps & Platform Engineering portfolio showcasing cloud infrastructure, Kubernetes, CI/CD automation, observability, and production-ready IaC with Terraform & Ansible
#automation#aws#ci-cd#docker#eks#grafana
Live repository statistics
★ 4Stars
⑂ 1Forks
◯ 0Open issues
◉ 4Watchers
57/100
OPENREPOHUB HEALTH SIGNAL
Mixed signals
A transparent discovery signal based on current public GitHub metadata.
Recent activity35% weight
72
Community adoption25% weight
9
Maintenance state20% weight
100
License clarity10% weight
0
Project information10% weight
100
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
README preview
██╗ ██╗██╗██╗ ██╗███████╗██╗ ██╗ ███████╗██╗███╗ ██╗ ██████╗ ██╗ ██╗
██║ ██║██║██║ ██║██╔════╝██║ ██╔╝ ██╔════╝██║████╗ ██║██╔════╝ ██║ ██║
██║ ██║██║██║ ██║█████╗ █████╔╝ ███████╗██║██╔██╗ ██║██║ ███╗███████║
╚██╗ ██╔╝██║╚██╗ ██╔╝██╔══╝ ██╔═██╗ ╚════██║██║██║╚██╗██║██║ ██║██╔══██║
╚████╔╝ ██║ ╚████╔╝ ███████╗██║ ██╗ ███████║██║██║ ╚████║╚██████╔╝██║ ██║
╚═══╝ ╚═╝ ╚═══╝ ╚══════╝╚═╝ ╚═╝ ╚══════╝╚═╝╚═╝ ╚═══╝ ╚═════╝ ╚═╝ ╚═╝
DevOps · MLOps · LLMOps Engineer
Building production-grade cloud-native infrastructure — from bare Terraform to LLM serving platforms
About
3+ years building and operating cloud-native infrastructure on AWS — across DevOps, MLOps, and LLMOps. I've shipped everything from Jenkins HA clusters and GitOps deployment platforms to self-hosted LLM inference gateways and automated ML retraining pipelines. Background in Python automation from high-volume VFX production (Disney+, Transformers, Ant-Man).
I care about zero-downtime deploys, full observability, and systems that run themselves.
Tech Stack
Cloud & IaC
AWS EKS · VPC · IAM · S3
RDS · ECR · EFS · ALB
Terraform Modules · Workspaces
Remote State
Ansible Roles · Packer AMIs
Containers & GitOps
Kubernetes EKS · Fargate · Helm
Docker Multi-stage builds
ArgoCD App-of-Apps · Rollouts
Kustomize Overlay environments
GitHub Actions OIDC · Matrix builds
ML / LLM Ops
MLflow Tracking · Registry
Kubeflow Pipelines v2
vLLM GPU inference
FastAPI OpenAI-compatible API
Evidently Drift detection
Observability
Prometheus Custom metrics · HPA
Grafana Dashboards · Alerting
Loki Log aggregation
Alertmanager PagerDuty · Slack
Languages
Python boto3 · asyncio · Pydantic
Bash Automation scripts
HCL Terraform modules
YAML K8s · CI/CD manifests
Platforms
Backstage IDP · Self-service
Redis Rate limiting · Cache
Jenkins HA on AWS ASG+EFS
Next.js Frontend deployment
Projects
🏗️ Infrastructure & Cloud
| Project | Highlights | |
|---|---|---|
| 01 | Jenkins HA on AWS | Terraform + Ansible + Packer · ALB → ASG EC2s → EFS shared home · Zero-downtime controller restarts |
| 03 | Scalable Java App on AWS | ALB → ASG → RDS Multi-AZ · Ansible provisioning · Auto-scaling on CPU |
| 05 | AWS VPC Design & Automation | 3-tier VPC (public / private / DB subnets) · NAT gateways · Security group layering |
📊 Observability
| Project | Highlights | |
|---|---|---|
| 04 | Prometheus Observability Stack | Prometheus + Grafana + Loki + Alertmanager · Docker + Terraform deploy · Slack alerting |
☸️ Kubernetes & GitOps
| Project | Highlights | |
|---|---|---|
| 08 | Fargate App Deployment | Serverless EKS Fargate · Helm chart · ALB Ingress Controller · No node management |
| 11 | EKS GitOps with ArgoCD | GitHub Actions CI → ECR → ArgoCD CD → EKS · Full deploy under 8 min · Drift detection |
🤖 MLOps
| Project | Highlights | |
|---|---|---|
| 15 | End-to-End MLOps Pipeline on AWS | Kubeflow Pipelines v2 · MLflow registry · ArgoCD GitOps promotion · Argo Rollouts canary · Evidently drift detection → auto-retraining · Model in prod < 15 min from merge |
S3 (versioned data)
└─► Kubeflow Pipeline: ingest → features → train → evaluate (F1 gate) → MLflow register
│
ArgoCD GitOps commit
│
EKS canary rollout (10% → 100%)
│
Evidently drift CronJob (every 6h)
└─► auto-trigger retraining
🧠 LLMOps
| Project | Highlights | |
|---|---|---|
| 14 | LLM Inference Gateway & Serving Platform | vLLM on EKS · OpenAI-compatible FastAPI · JWT (RS256) auth · Redis rate limiting · Custom metric HPA · p95 latency reduced 40% |
Client
└─► FastAPI Gateway
├─► JWT validation (RS256)
├─► Redis token-bucket rate limiter (free / standard / premium)
└─► Model router
├─► vLLM (Llama-3-8B) GPU node pool
├─► vLLM (Mistral-7B) GPU node pool
└─► vLLM (CodeLlama-13B) GPU node pool
│
Prometheus: token_throughput · TTFT · queue_depth
HPA: scale on vllm_queue_depth > 5
🛠️ Developer Platform & Automation
| Project | Highlights | |
|---|---|---|
| 09 | GitHub Actions OIDC → AWS | No long-lived credentials · OIDC JWT → STS AssumeRole → scoped IAM role |
| 12 | Backstage IDP Templates | Self-service service scaffolding · Auto-creates GitHub repo + ArgoCD app + CI pipeline |
| 13 | Platform Automation Scripts | boto3 + kubernetes-client · Stale resource cleanup · ECR image GC · K8s health reports → Slack |
Key Results
🚀 Full deploy cycle under 8 min EKS GitOps platform (zero manual steps)
⚡ p95 LLM latency reduced 40% vLLM batching + HPA tuning
🤖 Model to production < 15 min MLOps pipeline (merge → EKS canary)
🔄 Auto drift-triggered retraining Evidently AI + Kubeflow Pipelines
🔐 Zero long-lived AWS credentials GitHub Actions OIDC across all CI/CD
Work Experience
Neevya Therapy (2024 – Present) — DevOps Engineer (Freelance)
Sole engineer for an AI therapy chatbot platform. Built entire AWS infrastructure (EKS, VPC, IAM, RDS, S3) from scratch with Terraform. CI/CD via GitHub Actions + Argo Rollouts canary. Full Prometheus/Grafana observability stack.
Harikatha Studio (11/2024 – 01/2026) — Pipeline & Automation Engineer
Houdini → Unreal Engine bridge. FastAPI microservices exposing studio tooling as REST APIs. Procedural asset generation workflows.
Technicolor Animation Productions (07/2023 – 06/2024) — FX Artist TD
Python automation across high-volume VFX pipelines. FastAPI pipeline APIs. Productions: Crater (Disney+), Transformers: Rise of the Beasts, Ant-Man and the Wasp: Quantumania.
All projects include Terraform IaC, Kubernetes manifests, CI/CD pipelines, and observability config. Each folder is self-contained with a README covering architecture, quick-start, and design decisions.