Roshan Kr. Singh·Nov 8You Approve 97% of Your AI Agent’s Permission Prompts. Anthropic Said So Themselves.And somewhere in the 3% you don’t review, an agent is deleting 2.5 years of backups in a single command.
Roshan Kr. Singh·Oct 14Why I Had to Unlearn Everything Before I Could Learn AI as a DevOps Engineer(Subtitle: I didn’t need another AI course. I needed someone to tell me why Bedrock exists at all.)
Roshan Kr. Singh·Jun 5DevOps Isn’t Going Away — The Stack Is Just Getting TallerA few years ago, things felt simpler.
Roshan Kr. Singh·May 21Your LLM Gateway Works. But Do You Know What Each Call Costs?The next thing every AI gateway needs after auth.
Roshan Kr. Singh·Apr 23Why Your AI API Needs Authentication and Observability (Not Just a Model Call)The mistake most of us make when starting with LLMs
Roshan Kr. Singh·Apr 10How I Actually Learned AI + DevOps by Building a Real System (Not Just Watching Tutorials)When I started learning AI, I followed the usual path.
Roshan Kr. Singh·Mar 24Building a secure LLM gateway on AWS using API Gateway, Lambda, and Amazon BedrockLarge language models are easy to call. Most examples look like this:
Roshan Kr. Singh·Feb 27DevOps to AIOps: Why Traditional Monitoring Is Breaking (and What Comes Next)If you’ve worked in DevOps or SRE for a few years, this probably sounds familiar:
Roshan Kr. Singh·Feb 12Why Logs Are the Worst Place to Start DebuggingIf you’ve been in DevOps or SRE long enough, you’ve heard this sentence:
Roshan Kr. Singh·Feb 5RAG Is Not the Brain — Why Databases Don’t Fix Incidents in ProductionIf you’ve worked in DevOps or SRE for a few years, you already know this uncomfortable truth: