– Education
Diploma in Computer Engineering
Thiagarajar Polytechnic College, Salem
Co-founder profile
Karthick S
Co-founder, MiniAI Labs
- DevOps
- Cloud Infrastructure
- SRE
- AI/LLMOps Engineer
Cloud infrastructure, CI/CD and the reliability work that keeps production systems running. Karthick S is co-founder of MiniAI Labs and works from Chennai, India.
- Based in
- Chennai, India
- Experience
- Four years of DevOps and site reliability engineering
- Latest role
- Software Engineer – DevOps, HCL Technologies
- Education
- B.E. Computer Science, Sona College of Technology
Profile
An engineer who keeps production running
DevOps and site reliability engineer who designs cloud infrastructure on AWS and Azure, builds CI/CD pipelines, runs containerised services on Kubernetes and sets up the monitoring, alerting and incident response around them, most recently on production banking systems in a regulated, high-availability environment. He also applies LLMOps to operations work, with AWS Bedrock, Claude, GPT and LangChain.
The work started in production support. From September 2022 Karthick worked at HCL Technologies in the banking domain for ANZ Banking Group, a global bank operating across Australia, New Zealand and international markets. He watched performance and service levels across its banking services, handled P1 and P2 incidents end to end as part of the on-call rotation, and ran weekly health checks and compliance audits against ITIL practice.
From November 2023 he moved into DevOps on the same production applications: reliability against defined SLOs with Prometheus and Grafana alerting, CI/CD pipelines in Jenkins and GitLab across development, staging and production, containerised microservices on Docker and Kubernetes with rolling deployments designed to avoid downtime, and AWS provisioning with Terraform so every environment stays identical. Cloud cost was part of the job too, from EC2 rightsizing to S3 lifecycle policies.
Operations is also where he applies AI. He put an LLMOps assistant on AWS Bedrock into the on-call workflow to summarise logs during troubleshooting, and he works with OpenAI GPT, LangChain and n8n for automation. As co-founder of MiniAI Labs, his discipline is the infrastructure side of AI systems: where they run, how they ship and how they are watched once they are live.
Reliability
How production stays up
The path a change takes through the pipelines Karthick builds, drawn as one 3D illustration: a release passes five gates, Kubernetes replaces pods one at a time, and monitoring watches every pod while it happens. The three panels below explain the same steps without the animation.
A change travels a build track through five gates, then replaces pods one at a time across a three-node cluster while a monitoring ring above it flags one pod, which is replaced before the rollout continues.
Ship through five gates
Every release follows the same track, and a change that fails a gate never reaches the cluster. Small artefacts move along it continuously, one after another, rather than in one large batch.
Commit
A change lands in Git and starts a pipeline run. With GitOps tools such as ArgoCD or Flux, the repository is the record of what should be running.
Build
Jenkins, GitLab CI/CD or GitHub Actions build one versioned Docker image, and the same image moves through development, staging and production.
Test
The automated checks a release has to pass run against that image before it can move on, so a failing change stops here instead of in production.
Scan
Security scanning with tools such as Trivy, SonarQube and Snyk looks at the image and the code for known issues before anything is deployed.
Deploy
Helm or ArgoCD hands the new version to Kubernetes, which starts the rolling update shown on the cluster.
Roll out one pod at a time
Kubernetes (EKS, AKS or GKE) replaces the running pods across three nodes one at a time, one pod per node in turn, so no node loses more than one pod at once and the service keeps answering during the release. These are the rolling deployments he ran on production banking applications.
Watch every pod
Prometheus and Grafana watch the cluster against SLOs and error budgets. When one pod raises an alert, the signal on the ring spikes, the pod is replaced and the rollout carries on; afterwards a blameless postmortem records the root cause, as in his on-call work.
An illustration of how he works, drawn from his CV. It is not a client system and it shows no live data.
Career
Career, in dated steps
Education first, then each role with its dates and what the work involved, ending with MiniAI Labs. Employer names are his own career history, as listed on his LinkedIn profile; the outcome numbers on his CV are left out because they cannot be checked from here.
– Education
B.E. Computer Science
Sona College of Technology, Salem
– Role
Associate Software Engineer
HCL Technologies, banking domain (ANZ Banking Group)
- Technical support for a global bank operating across Australia, New Zealand and international markets
- Performance and service-level monitoring across banking services, catching bottlenecks before customers felt them
- P1 and P2 incidents handled end to end as part of the on-call rotation
- Weekly health checks and compliance audits against ITIL practice and ISO standards
- Knowledge-base articles for recurring issues
– Role
Software Engineer – DevOps
HCL Technologies, banking domain (ANZ Banking Group)
- Reliability for production banking applications against defined SLOs, with Prometheus and Grafana alerting and error-budget priorities
- CI/CD pipelines in Jenkins and GitLab CI/CD across development, staging and production
- Containerised microservices on Docker and Kubernetes (EKS) with rolling deployments designed to avoid downtime
- AWS provisioning with Terraform, designed to keep every environment identical and free of configuration drift
- Cloud cost work: EC2 rightsizing, S3 lifecycle policies and Lambda tuning
- Centralised monitoring and alerting with Grafana, Splunk and CloudWatch for on-call response
- Root-cause analysis and blameless postmortems for critical production incidents
- An LLMOps operations assistant on AWS Bedrock (Claude) in the on-call workflow, summarising logs to speed up troubleshooting
- Architecture, runbook and incident-response documentation
– presentCo-founder
Co-founder
MiniAI Labs
Builds
Four systems from his track record
Systems Karthick designed and built, described by what each one does and the tools it uses. They come from his CV: they are not MiniAI Labs client projects, and nobody they were built for is named.
- Built by Karthick
AI operations assistant on AWS Bedrock
An assistant on AWS Bedrock (Claude 3) that summarises CloudWatch logs, drafts deployment documentation and answers engineers’ Kubernetes and Terraform questions, wired into Slack and AWS services.
Stack
- AWS Bedrock (Claude 3)
- CloudWatch
- Slack
- Kubernetes
- Terraform
- Built by Karthick
Multi-cloud automation on n8n
Workflow automation connecting AWS EventBridge, Slack, GitHub and Azure APIs, so routine operational tasks run as monitored workflows instead of manual steps.
Stack
- n8n
- AWS EventBridge
- Slack
- GitHub
- Azure APIs
- Built by Karthick
Multi-tenant AI automation platform
A multi-tenant automation system on AWS using n8n and AWS Bedrock, with LLM-driven decisions, each tenant’s data kept separate by design, and the deployment, monitoring and reliability of the AI pipeline in production.
Stack
- AWS
- n8n
- AWS Bedrock
- Monitoring
- Built by Karthick
AI voice agent platform
Voice agents on Retell AI, ElevenLabs and Vapi AI that answer customer queries in conversation instead of a menu-driven phone system.
Stack
- Retell AI
- ElevenLabs
- Vapi AI
Skills
Eight domains, one operations stack
The tools he works with, grouped the way operations work happens: where systems run, how they ship, how they are packaged, watched and secured, and where AI fits in. Choose a domain to see its tools.
Cloud
Tools and methods
- AWS (EC2, Lambda, S3, IAM, VPC, EKS, ECS, RDS, CloudFront, CloudWatch, EventBridge)
- Azure (VMs, Functions, App Services, AKS, DevOps)
- GCP (Compute Engine, GKE, Cloud Run, Cloud Build)
CI/CD and GitOps
Tools and methods
- Jenkins
- GitLab CI/CD
- GitHub Actions
- Azure DevOps Pipelines
- AWS CodePipeline
- ArgoCD
- Flux
Infrastructure as code
Tools and methods
- Terraform
- Terragrunt
- CloudFormation
- Pulumi
- Ansible
- Packer
- Helm
- Kustomize
Containers
Tools and methods
- Docker
- Kubernetes (EKS, AKS, GKE)
- Karpenter
- HPA / VPA
- Istio
- Linkerd
- NGINX and Traefik ingress
SRE and observability
Tools and methods
- Prometheus
- Grafana
- Datadog
- Splunk
- ELK / EFK
- Loki
- Jaeger
- OpenTelemetry
- PagerDuty
- SLIs, SLOs and error budgets
- Postmortems
DevSecOps
Tools and methods
- HashiCorp Vault
- AWS Secrets Manager
- IAM and RBAC
- Trivy
- SonarQube
- Snyk
- Falco
- SAST / DAST
- OWASP
AI and LLMOps
Tools and methods
- AWS Bedrock
- OpenAI GPT
- LangChain
- Prompt engineering
- n8n
- Retell AI
- ElevenLabs
- Vapi AI
Data and systems
Tools and methods
- MySQL
- PostgreSQL
- DynamoDB
- Redis
- S3
- Python
- Bash
- Linux (Ubuntu, CentOS, RHEL)
- Windows Server
Education
Education and training
Education
- –Diploma in Computer EngineeringThiagarajar Polytechnic College, Salem
- –B.E. Computer ScienceSona College of Technology, Salem
Training
- UdemyTerraform & Infrastructure as Code (IaC)
- UdemyDocker & Kubernetes: The Complete Guide
- SimplilearnDevOps Certification Training – CI/CD, Docker, Kubernetes & Terraform
Course completions, not vendor certifications, and not certifications of MiniAI Labs.
Focus
Where his work sits
The six areas his CV and LinkedIn profile centre on, from cloud platforms to the LLMOps work that brings AI into operations.
- Cloud infrastructure (AWS, Azure, GCP)
- CI/CD and GitOps
- Kubernetes and containers
- Monitoring and incident response
- DevSecOps
- AI automation and LLMOps
FAQ
Questions about Karthick
What is Karthick S’s role at MiniAI Labs?
He is co-founder. His own discipline is DevOps and site reliability engineering: cloud infrastructure on AWS and Azure, CI/CD pipelines, Kubernetes, and the monitoring, alerting and incident response around production systems. This page sets out his background; how a project is scoped and delivered is described on the engagement models page.
Where did Karthick work before co-founding MiniAI Labs?
At HCL Technologies, in the banking domain for ANZ Banking Group, from September 2022 to August 2026. He started as an Associate Software Engineer on monitoring and P1 and P2 incident handling, and from November 2023 worked as a Software Engineer in DevOps on CI/CD, Kubernetes, Terraform and observability for production banking applications.
What does the 3D scene on this page show?
An illustration of the release path he works on, not a client system and not live data. A change passes five gates (commit, build, test, scan and deploy), Kubernetes then replaces eighteen pods on three nodes one at a time, and a monitoring ring above the cluster flags one pod, which is replaced before the rollout continues.
Which cloud platforms and tools does Karthick work with?
Mainly AWS (EC2, EKS, Lambda, S3, IAM, VPC, CloudWatch and more) and Azure (AKS, App Services, Functions, Azure DevOps), with GCP (GKE, Cloud Run) as well. Around them: Terraform and Ansible for infrastructure as code, Jenkins, GitLab CI/CD and ArgoCD for delivery, and Prometheus, Grafana, Splunk and Datadog for observability.
How does Karthick use AI in operations work?
As LLMOps inside the on-call workflow. He built an operations assistant on AWS Bedrock with Claude that summarises CloudWatch logs to speed up troubleshooting, drafts deployment documentation and answers engineers’ Kubernetes and Terraform questions, and he automates routine operational tasks as monitored n8n workflows.
Are Karthick’s courses vendor certifications?
No. They are training courses: Terraform and infrastructure as code, and Docker and Kubernetes, on Udemy, plus DevOps training on Simplilearn that covers CI/CD, Docker, Kubernetes and Terraform. They are listed as training, not as vendor certifications, and they are not certifications of MiniAI Labs.
Are the four builds on this page MiniAI Labs client projects?
No. They come from Karthick’s CV and are part of his own track record, described by what each system does and the tools it uses. Nobody they were built for is named, and none of them is presented as MiniAI Labs client work or as a system running on this site.
Next step
Plan an AI system that stays up after launch.
Book a free 30-minute strategy call. Bring one workflow; you will leave with a view on scope, the systems involved and how it would be deployed and monitored.
Free 30-minute call · no commitment