Keeping production boring since 2009.
I design production-ready systems, build the monitoring that tells you what is actually happening, and run the CI/CD pipelines that ship it. 16 years in production, from 24/7 incident rotations to leading SRE transformations on cloud-native platforms. Today, AI agents like Claude, Codex, Gemini and Cursor sit in my daily loop for triage, infrastructure code and postmortems.
Rajesh Dhar
Lead Site Reliability Engineer (SRE). I help startups fix the monitoring, reliability and cloud-cost problems that only show up once they're already in production.
- in production
- 17y 0m
- Nov 2009 → now
- services
- 6
- employers
- deploys
- 10
- role changes
- artifacts
- 6
- certifications
- agents
- 4
- AI tools in the loop
Current role: Lead Site Reliability Engineer since 2021.
Career trace
GET /career · select a span for its attributestrace_id rd-2009-11spans 15status OK
span
201020122014201620182020202220242026
durcurrent-rolefinastramanulifeopentextearly careereducation (hatched)
AI in the loop
kubectl get pods -n aiagents.yaml4 agents · human-approved applies
claude-codeagent
vendor: Anthropic- Terraform and pipeline changes as reviewed PRs
- Runbook and postmortem drafts from incident timelines
codexagent
vendor: OpenAI- Scripted toil: migrations, config sweeps, test scaffolds
- Second-opinion review on risky diffs
geminianalyst
vendor: Google- Long-context reads over logs, traces and dashboards
- Summaries of noisy alert history
cursoreditor
vendor: Anysphere- Day-to-day IaC, Helm and scripting
- Fast navigation of unfamiliar repos during incidents
policyAgents propose; a human approves every plan and apply.No production credentials in agent sessions.Every AI-assisted change ships through the same CI checks as mine.
Event log
tail -f /var/log/career.log39 events
nowINFOcurrent-roleagents in the loop: claude-code, codex, gemini, cursor; humans approve every plan and apply
nowINFOcurrent-rolemulti-cloud estate: aws, gcp, azure; provisioned with terraform + terragrunt
nowINFOcurrent-rolepipelines: jenkins, github-actions; observability: datadog; security compliance end to end
2021-10DEPLOYcurrent-rolerole=lead-sre started; scope="CI/CD, monitoring, security compliance"
2021-10INFOcurrent-roledesign and deliver production-ready solutions to client needs, with AI-driven insights
2021-10INFOcurrent-roleintegrating AI tooling into delivery pipelines where it earns its place
2021-10INFOcurrent-rolelead projects toward stable, reliable environments; own escalations from the service desk
2021-10INFOcurrent-roleglobal team: devops automation, knowledge sharing with dev teams, tight deadlines met
2021-10INFOcurrent-roletranslate complex technical trade-offs for non-technical stakeholders
2021-04DEPLOYopentextrole=lead-reliability-admin; on-call=24/7/365; itil=true
2021-04INFOopentextaws: ec2 rds vpc iam route53 s3; incidents resolved against SLA with full ownership
2021-04INFOopentextdiagnosed outages in high-throughput web applications and network services
2021-04INFOopentexttechnical liaison to product dev, customer service and sales for bug triage
2021-04INFOopentextmaintained and scaled configuration management across the stack
2020-09DEPLOYfinastrarole=expert-sre; design_authority="SRE patterns & practices"
2020-09INFOfinastraled SRE process and technical management of cloud-native financial platforms
2020-09INFOfinastrachange management lead for the SRE transformation program
2020-09INFOfinastrareviewed SLAs on live production deployments; ran weekly operating reviews
2020-09WARNfinastraon-call toil capped at 25%; blameless postmortems for high-severity incidents
2020-09INFOfinastraresiliency improvements shipped; design reviews to keep delivered code resilient
2019INFOuwaterloomaster of management science, management of technology: completed alongside full-time work
2019-04DEPLOYmanuliferole=platform-engineer; platform=pivotal-cloud-foundry,azure(aks)
2019-04INFOmanulifeiac=terraform,chef; packaging=helm; secrets=hashicorp-vault
2019-04INFOmanulifeci=jenkins,github-actions; quality=sonarqube; artifacts=artifactory
2019-04INFOmanulifeshipped in java, python, node.js and ruby with tdd and continuous integration
2019-04INFOmanulifemonitoring: new-relic grafana dynatrace ca-apm; dashboards for support teams
2019-01DEPLOYmanuliferole=platform-reliability-engineer; apm instrumentation, health dashboards, monitoring roadmap
2018-01DEPLOYmanuliferole=apm-specialist; built "single pane of glass" across monitoring tools
2018-01INFOmanuliferolled out agent-based, agentless, synthetic and scripted monitoring
2018-01INFOmanulifeadded machine-learning analysis to monitoring and diagnostics
2017INFOudemycertified: machine learning with python and r
2016-10DEPLOYopentextrole=principal-analyst; mentoring, brown-bags, white papers, tech talks
2016-10INFOopentextbuilt and ran vmware lab environments across data centres
2016-10INFOopentextcross-team, cross-product escalations with leads and management
2012-10DEPLOYopentextrole=senior-analyst; enterprise incident ownership begins
2012-10INFOopentextprimary contact for corporate customers through every incident to resolution
2010-06DEPLOYwiprorole=technical-support-associate; british telecom clients
2009-11DEPLOY3i-infotechrole=it-infrastructure-consultant; onsite client engagement
2009-11INFO3i-infotechdowntime reduced 83% for onsite client; bottlenecks removed in real time
What I run
kubectl get pods -A| NAMESPACE | NAME | STATUS | LAST SEEN AT |
|---|---|---|---|
| ai | claude / claude-code | Running | daily |
| ai | openai-codex | Running | daily |
| ai | gemini | Running | daily |
| ai | cursor | Running | daily |
| observability | datadog | Running | current-role |
| observability | new-relic | Running | manulife |
| observability | grafana | Running | manulife |
| observability | dynatrace | Running | manulife |
| observability | prometheus | Running | certified |
| observability | ca-apm | Running | manulife |
| cloud | aws (ec2, rds, vpc, iam, route53, s3) | Running | current-role |
| cloud | gcp | Running | current-role |
| cloud | azure | Running | current-role, manulife |
| cloud | aks / kubernetes | Running | manulife |
| cloud | pivotal-cloud-foundry | Running | manulife |
| iac | terraform / terragrunt | Running | current-role |
| iac | chef | Running | manulife |
| iac | helm | Running | manulife |
| iac | hashicorp-vault | Running | manulife |
| ci-cd | ci/cd pipelines | Running | current-role |
| ci-cd | jenkins | Running | current-role, manulife |
| ci-cd | github-actions | Running | current-role, manulife |
| ci-cd | sonarqube / artifactory | Running | current-role, manulife |
| practice | slos & slas | Running | current-role, finastra |
| practice | blameless postmortems | Running | current-role, finastra |
| practice | incident mgmt (itil) | Running | current-role, opentext |
artifact registrycertifications & degrees
- sre/sla-and-sloVERIFIEDSite Reliability Engineering: Service-Level Agreements and Objectives
- k8s/monitoring-with-prometheusVERIFIEDKubernetes: Monitoring with Prometheus
- terraform/network-infrastructureVERIFIEDTerraform: Managing Network Infrastructure
- ci/github-actionsVERIFIEDLearning GitHub Actions
- ml/python-and-r:2017VERIFIEDMachine Learning with Python and R · Udemy
- data/power-bi-with-excelVERIFIEDUsing Power BI with Excel
- uwaterloo/mmsc-mot:2019DEGREEMaster of Management Science, Management of Technology · University of Waterloo
- centennial/business-ops:2012DEGREECollege Diploma, Business Operations · Centennial College
- iiit-kolkata/be-it:2009DEGREEBachelor of Engineering, Information Technology · IIIT Kolkata
Work with me
pipeline: engagement · 4 stages · stage 0 is freeproof: 1,300+ datadog monitors · 116+ production services · slo/error-budget owner · dr lead · sre design authority, finastra
stage 0 · diagnose
Free audit
Send your anonymized AWS bill, or give me read-only access to your DataDog. You get a written report: the three things I'd fix first. Yours to keep either way.
- cloud-costyour anonymized AWS bill, analyzed: the top three places you're overpaying
- observabilitya read-only review of your DataDog: what's noisy, what's missing, the alerts that would have caught your last incident
asyncwritten report
freeno call required
get the free auditstage 1 · assess
Observability audit
Two weeks on your monitoring and alerting. You get a written map of blind spots and a ranked fix list.
coverage-scan2 wks
reportfixed price
stage 2 · build
Infrastructure project
Terraform modules, CI/CD pipelines, Kubernetes migration or alerting setup, scoped and priced up front.
terraform apply4–8 wks
handover + runbooksfixed scope
stage 3 · operate
Fractional SRE
A few hours a month on reliability: SLOs, PR reviews, incident support and postmortems.
monthly review15–20 h/mo
on-call escalationrolling
stage 0 is free. stages 1–3 are the paid work.
Before you page me
- nda?
- Yes, on request, before any access.
- how does stage 0 work?
- You send an anonymized bill or read-only Datadog access. I send a written report. No call required, no obligation.
- where are you?
- Kitchener, Ontario. I work in your timezone's core hours.
- who does the work?
- Me. No hand-offs, no offshore team.
- what do you need from us?
- One person who can answer access questions. Nothing else.
PHP2Shell: patch what you need, delete what you don't
Two critical PHP flaws landed on a Tuesday. One service got patched within the day. A handful of old PHP scripts got something better than a patch: they got switched off for good.
An AI agent's plan that looked right and wasn't
AI agents are in my daily loop for infrastructure code. One produced a tidy change whose plan would have replaced a load balancer. The guardrails are why it didn't.
1 min readA generated config file that doubled in size
What I took from Cloudflare's November 2025 outage: a permissions change made a generated file double in size, and the proxy crashed on a limit nobody expected to hit.
1 min read
Page me
opens a ticket in my inbox, not a 3am alertTell me what is breaking, what is slow, or where AI agents could take toil off your team. I reply within a day.
open incident