Shaswat Raj · DevOps / Platform Engineer
I care about what happens after the merge: deploys, failure modes and observability.
My infrastructure experience comes from two places: contributing to cloud-native projects (tracing and service mesh), and deploying and operating my own services on edge and container platforms with CI/CD that I wrote.
Looking for: DevOps, platform, SRE and infrastructure internships or new-grad roles.
Evidence
What I have done that is relevant
jaegertracing/jaeger #10024 — sampling strategy validation
Rejects out-of-range parameters in file-based sampling strategies instead of accepting a config that silently misbehaves. Go, in the CNCF tracing project.
in reviewkmesh-net/kmesh #1977 and #1983 — service-mesh dataplane tooling
Decide command success by exit status rather than stderr output, and match the dataplane-mode label the same way the daemon does. Go.
in reviewCI/CD for this site — build, deploy and mirror on every push
GitHub Actions builds the static export and deploys it to Pages with scoped permissions and a concurrency group. A second workflow mirrors main to GitLab, which runs its own Pages pipeline.
liveServices in production on Cloudflare Workers
IndexFast, MCP Pure and Solus run on Workers with Hono, Neon Postgres, Drizzle and R2 — deployed and operated by me.
liveBitotsav '26 — operating a system during a live event
Redis in front of PostgreSQL for hot paths, entry scanning that degrades gracefully without network, and on-call through the festival.
liveNetScopeX — observability platform
A metrics and traces platform in strict TypeScript on PostgreSQL, built to understand the reliability tooling I contribute to.
in progress
Tools I have shipped with
How I work
- Fail loudly on bad config instead of running with it.
- Automate the deploy before the second deploy.
- Assume the network will fail and decide what the system does then.