I am a DevOps Engineer and AI/ML Infrastructure specialist with production experience across Kubernetes, Docker, CI/CD, GitOps, Infrastructure as Code, and observability. I work on the intersection of platform engineering, SRE, and MLOps: deploying reliable cloud-native systems, automating delivery pipelines, and making AI-powered tools production ready.

My recent work includes operating workloads across AKS, EKS, and GCP, building SLO-aware observability with Prometheus, Grafana, Loki, Alertmanager, and OpenTelemetry, and deploying AI/LLM systems such as a Kubernetes SRE Agent using RAG, Gemini API, OpenAI API, Redis caching, and GitOps.

name: Nishit Ranjan Saha
role: DevOps Engineer | AI/ML Infrastructure | MLOps Engineer
company: TCG Digital
location: Kolkata, India
experience:
  - Product Developer, TCG Digital
  - Consultant, TCG Digital
strengths:
  - Kubernetes production operations
  - CI/CD and GitOps automation
  - MLOps and LLM-powered infrastructure tooling
  - Observability, SLOs, incident response, RCA
  - Infrastructure as Code and secure deployments

June 2025 - Present | Kolkata, India

• Deployed and operated software across 10+ client environments on AKS, EKS, and GCP using Kubernetes, Docker, Helm, and Terraform.
• Maintained 99.9% uptime with SLA/SLO-compliant production operations.
• Architected the VorteX Observability Platform using Prometheus, Loki, Grafana, Alertmanager, and OpenTelemetry.
• Reduced MTTD by 40% with proactive monitoring, alerting, dashboards, and incident visibility.
• Designed Jenkins and GitHub Actions CI/CD pipelines with GitOps workflows using ArgoCD.
• Cut pipeline configuration time by 60% and standardized deployments across client environments.
• Implemented Ansible playbooks and Helm charts for repeatable one-click deployments of VorteX and Vault-KES.
• Reduced setup time from hours to minutes through Infrastructure as Code automation.
• Integrated HashiCorp Vault, MinIO, and KES for enterprise-grade secrets management.
• Hardened container images, performed CVE remediation, and supported release compliance.
• Managed 3 concurrent production environments with Kubernetes rolling deployments and autoscaling.
• Automated Jira reporting with Python, saving 5+ hours weekly and improving scrum velocity by 20%.

December 2024 - June 2025 | Kolkata, India

• Applied Docker, Kubernetes, Ansible, Linux, Python, Elasticsearch, and MySQL across enterprise client environments.
• Achieved 98% SLA compliance through systematic incident prioritization and queue management.
• Reduced average ticket resolution time by 20% with structured RCA documentation.
• Implemented Terraform and Ansible provisioning across 5+ client environments.
• Improved deployment efficiency by 25% and reduced configuration drift.
• Delivered GitOps deployment workflows adopted across multiple client engagements.
• Expanded Java and JavaScript delivery capability within 3 months of onboarding.

flowchart LR
    A["Developer Commit"] --> B["GitHub Actions / Jenkins"]
    B --> C["Build and Test"]
    C --> D["Container Image"]
    D --> E["Security Scan and CVE Checks"]
    E --> F["Helm Chart Update"]
    F --> G["ArgoCD Sync"]
    G --> H["Kubernetes Rollout"]
    H --> I["Prometheus and Grafana Validation"]
    I --> J["SLO / SLI Review"]

flowchart TD
    U["User Query"] --> API["Agent API"]
    API --> INTENT["Intent Detection"]
    INTENT --> POD["Pod Diagnostics"]
    INTENT --> NODE["Node Diagnostics"]
    INTENT --> SVC["Service Diagnostics"]
    POD --> RAG["RAG Context"]
    NODE --> RAG
    SVC --> RAG
    RAG --> CACHE{"Redis Cache Hit?"}
    CACHE -- "Yes" --> ANSWER["Response"]
    CACHE -- "No" --> LLM["Gemini / OpenAI API"]
    LLM --> ANSWER
    ANSWER --> OBS["Trace, Logs, Metrics"]

flowchart TB
    APP["Applications"] --> OTEL["OpenTelemetry"]
    APP --> PROM["Prometheus"]
    APP --> LOKI["Loki"]
    OTEL --> TRACE["Distributed Traces"]
    PROM --> GRAF["Grafana Dashboards"]
    LOKI --> GRAF
    TRACE --> GRAF
    PROM --> ALERT["Alertmanager"]
    ALERT --> INCIDENT["Incident Response"]
    INCIDENT --> RCA["RCA and Remediation"]

flowchart LR
    TF["Terraform"] --> AKS["Azure AKS"]
    TF --> EKS["AWS EKS"]
    TF --> GCP["GCP Workloads"]
    ANS["Ansible"] --> CONFIG["Configuration Management"]
    HELM["Helm"] --> RELEASE["Application Release"]
    CONFIG --> AKS
    CONFIG --> EKS
    CONFIG --> GCP
    RELEASE --> AKS
    RELEASE --> EKS
    RELEASE --> GCP

• Building production-ready AI infrastructure on Kubernetes.
• Improving GitOps deployment reliability across multi-cloud environments.
• Designing observability stacks that reduce MTTD and support fast RCA.
• Optimizing LLM applications with caching, batching, intent routing, and SLO-aware design.
• Hardening containers, remediating CVEs, and improving release compliance.

Automate repeatable work.
Make production behaviour observable.
Design for rollback before release.
Treat reliability as a feature.
Bridge AI capability with platform discipline.

Add your blog links here when ready. Suggested topics based on your CV:

• Building a Kubernetes SRE Agent with RAG and LLM intent routing.
• Designing a Prometheus, Loki, Grafana, Alertmanager, and OpenTelemetry platform.
• GitOps release patterns with Jenkins, GitHub Actions, ArgoCD, and Helm.
• Redis caching strategies for reducing LLM API cost and latency.
• CVE remediation and container image hardening in CI/CD pipelines.