---

$ whoami

$ kubectl get engineer amllan -o wide
NAME     ROLE              COMPANY   EXPERIENCE   STATUS    FOCUS
amllan   DevOps Engineer   Jar       2+ yrs       Running   K8s · AWS · GitOps · Observability · AIOps

$ kubectl describe engineer amllan | grep -A6 Spec
Spec:
  Builds:      Kubernetes platforms, GitOps pipelines, internal developer tooling
  Runs:        LLM gateway for 400+ developers (SSO, budgets, autoscaling)
  Believes:    toil should be automated, alerts should be actionable
  Learning:    SRE: SLOs, error budgets, incident response, Linux internals
  Building:    aiops-platform (from-scratch AIOps + MLOps on AWS)
  Location:    India 🇮🇳

⚡ Highlights

☸️ Kubernetes platform Node autoscaling with Karpenter and workload scaling with HPA Per-team staging namespaces for isolated testing Safe rollouts with health probes and quick rollbacks	🔁 GitOps and developer platform Argo CD for declarative, Git-driven deployments Backstage self-service portal for developers Centralised secrets with HashiCorp Vault
🤖 AI platform Runs an internal LLM gateway (Bifrost) for 400+ developers SSO login, per-user budgets, and autoscaling (HPA 20 → 50 pods) Cost and usage visibility per team	💬 AI support bot A CRM assistant answering support questions with an LLM Guardrails scope every answer to the signed-in user's own data Blocks system-level and cross-user questions

🛠️ Now building: aiops-platform

A production-style AIOps + MLOps platform, built from zero, fully in Git: detect → explain → fix (with human approval).

flowchart LR
    subgraph K8s["☸️ Kubernetes on AWS (Terraform + Argo CD)"]
        APP["Demo app<br/>with chaos switches"]
        OBS["Prometheus · Loki<br/>OpenTelemetry"]
        AM["Alertmanager"]
    end
    subgraph AI["🤖 AIOps layer"]
        DET["Log + metric<br/>anomaly detection"]
        ENR["AI alert enricher"]
        REM["Remediation agent<br/>(approval-based)"]
    end
    GW["Bifrost LLM gateway"] --> BR["Amazon Bedrock<br/>Claude · Titan"]
    APP --> OBS --> AM --> ENR
    OBS --> DET --> ENR
    ENR <--> GW
    ENR --> SL["💬 Slack"]
    SL -- "approve" --> REM --> APP
    DET -. "models tracked in" .-> ML["MLflow · Evidently"]

Loading

Layer	Stack
Infra	Terraform · Kubernetes on AWS · GitHub Actions · Argo CD
Observability	Prometheus · Alertmanager · Grafana · Loki · OpenTelemetry
LLM gateway	Self-hosted Bifrost → Amazon Bedrock (Claude, Titan embeddings)
AIOps	AI alert enricher · log anomaly detection (Drain3) · approval-based remediation agent · LLM evals in CI
MLOps	MLflow registry · Evidently drift detection · automated retraining · shadow/canary model deploys

🧰 Tech stack

☁️ Cloud
📦 Containers & orchestration
🏗️ IaC, config & secrets
🔁 CI/CD & platform
📈 Observability
🐧 OS, scripting & tooling
🗄️ Data
🤖 AI & MLOps
🌐 Web (MERN)

📌 Featured repositories

Repository	What it is
🚨 Scoutflo-SRE-Playbooks	Incident-response playbooks for AWS and Kubernetes: step-by-step guides for on-call engineers
🧪 DevOps-Projects	Real-world DevOps projects, from beginner to advanced
📈 Microservice-Monitroing	Monitoring stack for a microservices application
⚙️ k8s-Ansible	Ansible automation to bootstrap a Kubernetes cluster
☁️ EKS_Terraform	Amazon EKS provisioned with Terraform

📊 GitHub stats

🐍 Contribution snake