Krishna Neupane

1. 01

   Security belongs in the workflow

   Controls run inside the pipeline on every change and fail closed. A check that depends on someone remembering it is applied inconsistently.

2. 02

   Infrastructure is reproducible

   Everything is declared in code, reviewed, and versioned. If an environment can't be rebuilt from Git, it can't be trusted.

3. 03

   Design for failure

   Multi-AZ layouts, tested restores, and runbooks are part of the design, not an afterthought, because something always fails eventually.

4. 04

   Platforms serve developers

   The secure, standard path should also be the easiest one. A platform succeeds when teams choose it, not when they're forced onto it.

5. 05

   Observability drives decisions

   Metrics, logs, and traces exist to answer questions during an incident and to show whether a change actually helped.

6. 06

   Keep it boring and measurable

   Prefer simple, maintainable systems with clear owners and numbers that show they work over clever ones nobody can operate.

This site follows the same rules: a signed SBOM, documented architecture decisions, and a published threat model. Verify it →

1. May 2024 – Present

   Parkcity Technologies

   Senior DevSecOps Engineer

   • 5–6 engineers led and managed
   • 10+ Kubernetes clusters
   • 300+ services in production
   • 10M+ end users on the platforms we run
   • 99.99% uptime
   • $84K+/yr cloud savings for one customer ($19K → $12K/mo)
   • v1.35 Kubernetes, upgraded from v1.33

   • —Lead and manage a team of 5–6 engineers, owning priorities, delivery, and quality from design to production.
   • —Operate 10+ production Kubernetes clusters running 300+ services at 99.99% uptime across AWS, GCP, and DigitalOcean, for customer products with more than 10 million users.
   • —Lead Kubernetes upgrades and cluster lifecycle management, most recently 1.33 → 1.35, along with database modernization.

2. Dec 2021 – May 2024

   BeyondID

   Okta Apex Partner

   DevSecOps Engineer

   • 5 environment tiers: dev, QA, UAT, MVP, prod
   • 4 clouds: AWS, GCP, DigitalOcean, Heroku

   • —Designed and automated infrastructure provisioning with Terraform and Terragrunt for secure, identity-focused services.
   • —Ran dev, QA, UAT, MVP, and production environments across AWS, GCP, DigitalOcean, and Heroku, plus Azure on one client project.
   • —Built and maintained Kubernetes deployment platforms for highly available production workloads.

Earlier: Software Engineer, DevOps at a software services company (2020–2021); System Engineer at a nationwide internet service provider (2017–2020).