Self-Healing Deployment
Trigger a fault on purpose, then watch the pipeline spot it and heal itself.
Instant verified answers about experience, US work status, 20 builds & skills.
Happy to answer anything about Srinivasan's work: his 5+ years of DevOps experience, US & India work authorization, 20 open-source builds, enterprise migrations, or contact details.
23 platform orchestrators, deployment tools, and cloud utilities. Designed to solve real infrastructure problems and tested in production.
A deployment pipeline that ships a new version of a web service, checks whether the new version actually works, and puts the old one back automatically when it does not. Two container slots sit behind one port: the outgoing version is stopped but kept, the new one takes the port, and a health gate decides which of them keeps it. Every attempt — success, rollback, or a build that never got that far — is written to a log that a small dashboard draws as a timeline. It generalises the rollback and validation automation I built for production microservices into an open demo you can run for yourself.
Trigger a fault on purpose, then watch the pipeline spot it and heal itself.
Drag-and-drop PDF processing that stays entirely in your browser. Your files stay yours.
A full recruitment-consultancy site on its own domain: services, reviews, contact.
One dashboard provisions free-tier compute across four clouds with Terraform, and an hourly sweep keeps every account tidy.
Claude does the planning, free-tier APIs do the legwork, and one request gets spread across five providers.
Point it at a codebase and it draws the map: a queryable graph of who calls what, plus chunks ready for a RAG pipeline.
Fifteen skills that teach Claude Code to spend fewer tokens on the same work, and two scripts that prove whether it actually worked.
Build an app through seven bounded phases — plan, test, implement, review, verify, remember, improve — each its own subagent.
A centralized Master Bot that supervises, secures, watermarks, and auto-triages every repository across the GitHub account.
An offline-first knowledge and education server: Wikipedia, courses, maps, local AI and RAG, with the internet unplugged.
A homelab chatbot that searches offline Wikipedia snapshots through a local LLM.
Reusable GitHub Actions templates that build, sign and release Windows and Linux installers.
Turns PDFs, Office files and scans into Markdown — as a GitHub Action on every push, or in your browser with nothing uploaded.
A WiX scaffold that turns a build output into a signed MSI, shortcuts and all.
Type in a topic and get back a finished short-form study video: script, voiceover, render.
A game server run entirely by editing text files: push to main, and a robot backs up, deploys and verifies.
Unified Raspberry Pi appliance pipeline: disk benchmarking, unattended multi-reboot bootstrapping, Matter Test Harness provisioning, and compressed image generation with auto-expansion.
Turns a bare Pi into a certified Matter test harness, then images the finished SD card.
The unattended bootstrap half of the imaging pipeline. It picks itself back up after every reboot.
One button in GitHub starts an Azure DevOps release, streams its logs back, and records the result.
Cuts a Raspberry Pi backup image down to the space actually used, and grows it back on first boot.
Index a photo folder onto a USB drive, then browse it in an Apple-Photos-style grid that runs entirely in the browser.
Sweeps dd block sizes from 512B to 64MB to find the fastest one for this disk.
Forks where I fixed a bug or extended behaviour upstream and sent the change back as a pull request.
A Minecraft server written from scratch in Rust — an open-source, drop-in alternative to the vanilla Java server, built for lower memory use and faster ticks.
The server enforced the per-player movement lock — used to hold a player still during dialogues and cutscenes — in the on-foot position handlers, but applied client-sent vehicle positions unconditionally, so a locked player could still drive a boat, minecart or horse anywhere. The patch rejects the packet and teleports the client back to the server-side position, mirroring what the on-foot handlers already do.
A private algorithmic-trading toolkit that scans market data and manages simulated positions. I contribute to it through a fork.
Reworked the market-data cache to source from the live-price feed and cut the intraday TTL from 60s to 15s, so scanners act on fresher quotes. Live exit-price lookups now try the real-time price first and fall back to the one-minute bar only when it is unavailable.