01 About
From hospitality
to SRE, to operations
that reason.
02 Expertise & selected work
What I'm good at,
and where it shipped.
Projects & automations
03 Journey
A 15-year
operating timeline.
From hotel reservation systems to AI-governed incident intelligence. Every role taught me how to make systems quieter and humans more leveraged.
04 AI Lab · AI in SRE
How I put models on-call
without putting trust at risk.
Patterns I've taken from idea to production. Each one exists because a simpler version failed first.
04.0 Case study
OIP — an Operational Intelligence
Platform, built from zero.
An AI-augmented operations platform for a payments on-call team. Designed, built and run by one engineer in six months — about two dozen repositories, a quarter-million lines, and a self-hosted triage model tuned from 30% to 75% in one pilot.
04.1 Architecture
Signals in, governed decisions out.
Simplified. Internal tool names are generalised; vendor names (Slack, Grafana, incident.io, Ollama, Trino) are real.
04.2 Cost-ordered inference
Spend the cheapest capable step first — and never let config reorder it.
Eligibility and health are tracked separately: a healthy model never gets promoted just because it is up. Models move through EXPERIMENTAL → SHADOW → HUMAN_REVIEW → QUALIFIED → APPROVED_PRIMARY, each step needing a named human and at least 30 adjudicated samples, with a regression gate of ±0.05 against the recorded baseline.
04.3 Measured, not claimed
A self-hosted model went from 30% to 75% agreement with human triage in a three-month pilot.
Held-out rows from a 7-day window. "Human label" is the team's manual triage category, used as a proxy — the honest caveat is that it is not blind adjudication. The qualification bench only ever runs on a copy of the database.
04.4 What's inside
Twenty-odd repositories, one operating model.
05 Value to a business
What this looks like
on your balance sheet.
Reliability work is only worth funding when it changes cost, risk or speed. This is where mine does.
Ways to work together
Hiring for a specific title? The role-fit page maps this experience onto Forward Deployed Engineer, AIOps, Staff SRE, platform architect and automation roles.
06 Skills & certifications
The toolbelt, honestly rated.
Years are hands-on production years, not "familiar with". Levels are self-assessed and I'm happy to be tested on any of them.
Certifications & courses
07 Knowledge
Mistakes I made,
and the architecture that fixed them.
Recent writing
08 Resume
One profile,
three formats.
International, GCC and India formats, generated fresh from the same source of truth that renders this page.
09 Contact
Let's build something
reliable.
Open to Senior Staff SRE, AI-operations, Forward Deployed Engineer and platform roles: remote for teams anywhere in the world, with relocation considered for the right role. Also open to advisory work. I reply within 24 hours.
tcgr@ops ~ %
$ whois chandraguptha
location: Bengaluru, IN · remote worldwide · relocation considered
role: Senior Staff Engineer — SRE @ Airbnb (via Altimetrik)
availability: 60 days notice (negotiable)
languages: English, Hindi, Tamil, Telugu, Kannada
timezone: IST (UTC+5:30) · overlaps GST, CET, US-East mornings
$