Devops or Platform Enginner Roadmap

1. Linux


Linux Architecture
Boot Process
Kernel
Systemd
Process Management
Memory Management
Disk Management
File System
Permissions
Networking
Logs
Package Management
Performance Troubleshooting


Commands to know: ls, cd, pwd, cat, grep, awk, sed, find, top, htop, df, du, free, tar, ssh, scp, systemctl


2. Networking


OSI Model
TCP/IP Model
TCP vs UDP
HTTP vs HTTPS
SSL/TLS
DNS
Load Balancer
Reverse Proxy
NAT
CIDR & Subnetting
VPN
Firewall
Routing


Important ports: 21, 22, 23, 80, 443, 3306, 5432, 8080, 53

Tools: ping, traceroute, nslookup, dig, netstat, ss, telnet


3. AWS


IAM (Users, Roles, Policies)
EC2 (Instances, AMI, Types)
VPC (Subnets, Route Table, IGW)
Security Groups vs NACL
Route53 (DNS, Routing Policies)
S3 (Storage Classes, Lifecycle)
EBS vs EFS
RDS (Multi-AZ, Read Replica)
ALB vs NLB
Auto Scaling
CloudWatch (Metrics, Alarms)
CloudTrail
Lambda
ECS vs EKS
Organizations & SCP
Transit Gateway
Direct Connect
KMS (Keys, Policies)



4. Docker


Docker Architecture
Docker Engine
Images vs Containers
Layers
Dockerfile (Instructions)
CMD vs ENTRYPOINT
COPY vs ADD
Volumes
Networking
Docker Compose
Multi-stage Builds
Docker Security
Docker Image Scanning
Troubleshooting


Docker commands: docker build, docker run, docker ps, docker exec, docker images, docker logs, docker compose


5. Kubernetes


Pods, ReplicaSet, Deployment
StatefulSet, DaemonSet
Job, CronJob
ConfigMap, Secret
Namespaces
Services (ClusterIP, NodePort, LoadBalancer, Headless)
Ingress
CoreDNS
PV, PVC, StorageClass, CSI
CNI
kube-proxy, kube-apiserver, etcd, kube-scheduler, kubelet


Security: RBAC, Service Account, IRSA, Network Policy

Auto Scaling: HPA, VPA, Cluster Autoscaler

Common Production Issues:


CrashLoopBackOff
ImagePullBackOff
OOMKilled
Pod Pending
Evicted Pods
Readiness Probe
Liveness Probe
DNS Failure
Pod Not Accessible
Error Codes
High Latency
Node NotReady
Exit Code 137
Debugging Steps



6. Terraform


Terraform Workflow: init, plan, apply, destroy
Providers (Local, Remote)
Backend & State
State Locking
Workspaces
Modules
Variables
Outputs
count vs for_each
Local-exec vs remote-exec
Drift Detection
Dependency Graph (DAG)
Reusable Modules
Infrastructure Rollback


Terraform commands: terraform init, terraform plan, terraform apply, terraform destroy, terraform refresh, terraform fmt, terraform validate


7. Ansible


What is Ansible?
Ansible Architecture
Inventory
Playbook
YAML Syntax
Ad-hoc Commands
Modules
Roles
Collections
Variables
Facts
Templates (Jinja2)
Handlers
Tags
Loops
Conditionals
Ansible Vault
Dynamic Inventory
Ansible Galaxy
Error Handling
Idempotency
Delegation


Common Questions:


How to restrict Nginx using Ansible?
How to manage secrets?
What is Command vs Shell?
Copy vs Template?
How to debug failed playbooks?
What is Idempotency?



8. CI/CD & Jenkins


Jenkins Architecture (Master-Agent)
Pipeline as Code (Jenkinsfile)
Stages, Steps
Agents (Static, Dynamic)
Build Triggers
Webhooks
Credentials Management
Shared Libraries
Parallel Pipeline
Rollback Strategy
Pipeline Failures
Security in Jenkins


Pipeline Stages: checkout → build → test → deploy → post actions


9. Git & GitHub


Git Basics
Branches
Merge vs Rebase
Cherry-pick
Reset vs Revert
Stash
Tags
Git Flow
Git Hooks
Pull Request
GitHub Actions (CI/CD)
GitOps


Important commands: git clone, git add, git commit, git push, git pull, git branch, git merge, git rebase


10. Observability & Monitoring


Monitoring vs Observability
3 Pillars (Metrics, Logs, Traces)
Prometheus
Grafana
Alertmanager
Node Exporter
OpenTelemetry


Tools: CloudWatch, Datadog, Alerting, SLO, SLA


11. Logging


ELK Stack (Elasticsearch)
EFK Stack (Fluentd/Fluentbit)
Logstash
Loki
Centralized Logging



12. Security


IAM & Least Privilege
RBAC in Kubernetes
Secrets Management
KMS (Key Management)
Encryption (At rest, In transit)
TLS/SSL Certificates
Image Scanning
Vulnerability Management
Application Firewall
Compliance



13. SRE


SLA, SLO, SLI
Error Budget
Burn Rate
Incident Management
RCA (Root Cause Analysis)
On-call
Disaster Recovery
Capacity Planning



14. Production Scenarios


Website returning 503
High CPU Usage
High Memory Usage
High Disk Usage
Database Slow
DNS Failure
SSL Certificate Expired
Terraform State Lock
ImagePullBackOff
Pod NotReady
Terraform Pipeline Failure
CrashLoopBackOff
Rollback Deployment
Pod Eviction
Application Down
Log Not Showing
Monitoring Alerts
Complete Outage Debugging



15. AI & Cloud


RAG (Retrieval Augmented Generation)
AWS Bedrock
Generative AI Basics
Use cases in DevOps

---

• Linux Mastery
• Linux Networking Mastery
• Linux Security Mastery
• Linux Observability Mastery
• Linux Performance Tuning Mastery
• Docker
• Ansible
• Git & GitHub Mastery
• Kubernetes Mastery
• Helm
• Terraform Mastery
• Jenkins Mastery
• DevOps Mastery
• Python for DevOps
• AWS Cloud Mastery
• Networking Mastery
• Bash Scripting Mastery
• Grafana
• Prometheus
• Loki
• ElasticSearch
• InfluxDB
• PostgreSQL
• MySQL
• JSON API

---

FOR AI/ML

If you're building the ultimate 2026 engineering library, I'd strongly recommend adding these as full standalone courses:

AI/ML Mastery (Mathematics → Deep Learning → Production AI)
LLM Engineering Mastery
AI Agents Mastery
MLOps Mastery
AI Infrastructure Engineering
RAG (Retrieval-Augmented Generation) Mastery
Vector Database Mastery
Prompt Engineering Mastery
Model Context Protocol (MCP) Mastery
AI Security & AI Safety Mastery
Generative AI Mastery
Computer Vision Mastery
Natural Language Processing (NLP) Mastery
Reinforcement Learning Mastery
Data Engineering Mastery
Big Data Engineering Mastery

---

AI/ML further Tech:

Agentic Systems
Orchestrated Workflows (loop)
RAG (Retrieval Augmented Generation)
MCP (Model Context Protocol)
MCP Server
MCP Clients
Tool Discovery

Agent Frameworks (LangGraph)
CrewAI(Role-based Teams)
Microsoft Agent Framework

Secure Tool Access

GraphRAG
Field-Theoretic Memory
Long-Context Management

LLMOps & Evaluation

Moving from "it works" to "it works reliably and cheaply."

Evaluation Frameworks: Tools like Arize Phoenix, LangSmith, or Ragas to measure hallucination rates, faithfulness, and answer relevance automatically.

Cost Optimization: Techniques like speculative decoding, model cascading (using small models for easy tasks, large for hard ones), and prompt caching.

Observability: Tracing full agent trajectories (not just single prompts) to debug why an agent got stuck in a loop or made a bad decision.

Local & Edge AI

Running models efficiently outside of big cloud APIs.

Inference Engines: vLLM (high throughput), Ollama (local dev), and MLX (Apple Silicon optimization).

Quantization: Techniques (GGUF, AWQ) to run large models on consumer hardware with minimal quality loss.

Small Language Models (SLMs): Fine-tuning and deploying models like Llama-3-8B or Phi-3 for specific, low-latency tasks.

---

Category    Key Technologies/Concepts   Why It Matters
Orchestration   LangGraph, CrewAI, Temporal Builds reliable, multi-step agent workflows. 
Connectivity    MCP, A2A Protocol   Standardizes tool access; prevents vendor lock-in. 
Memory  GraphRAG, Knowledge Graphs, Zep Enables complex reasoning over connected data. 
Ops LangSmith, Arize, Prometheus (AI metrics)   Ensures quality, tracks cost, and debugs agents.
Inference   vLLM, Ollama, Quantization  Reduces latency and cloud costs; enables privacy. 
Security    Guardrails AI, LLM Firewalls    Prevents prompt injection and data leakage. 

---

Advanced Penetration Testing & Exploitation

These tools move you from "running a scan" to "chaining vulnerabilities."

Burp Suite Professional: The industry standard for web app security. To gain depth, master Burp Extensions (writing your own in Java/Python), Intruder for complex fuzzing, and Repeater for manual request manipulation.

Metasploit Framework: Essential for understanding exploit development. Learn to write custom modules in Ruby, not just use existing ones, to understand buffer overflows and memory corruption.

Cobalt Strike: The standard for adversary emulation and red teaming. Learning its Beacon syntax and C2 (Command & Control) profiles teaches you how advanced persistent threats (APTs) operate undetected.

PentAGI / PentestGPT: New AI-driven agents that assist in chaining exploits. Learning to orchestrate these helps you understand how AI can automate complex attack paths, a critical skill for 2026 defense.