Gurugram, IN · open to senior SRE / platform roles · UK & remote, sponsorship
Prateek Rajvats
Platform & Reliability Engineer
I run the observability and reliability platform for 30+ engineering teams across a multi-account AWS org — turning compliance-critical infrastructure into paved roads that are safe by default.
About
Platforms as code, reliability as a habit
Platform engineer with 6+ years across DevOps and platform work, now leading observability and reliability at Safeguard Global for 30+ internal teams on a multi-account AWS organisation that runs payroll and workforce systems under SOC2 and CIS compliance. My work is the boring-in-a-good-way kind: self-service access without long-lived keys, monitors defined as code, supply-chain provenance checked before anything ships, and incidents that page the right team instead of everyone. I care about giving other engineers paved roads — and about the day the auditors show up going quietly.
Experience
Where I've worked
- Built a GitOps OIDC role-vending system — teams request scoped cross-account AWS access via PR-driven YAML, provisioned by Terraform, with no long-lived keys.
- Own the observability layer in Datadog: monitor-as-code for payroll Lambdas and SQS dead-letter queues, plus an OpsGenie → Datadog Incident Management migration.
- Shipped compliance-as-code: Checkov IaC scanning mapped to SOC2 / CIS controls, and an SBOM governance programme on JFrog Xray with CycloneDX.
- Designed EventBridge cross-account replication with tested bidirectional loop-prevention, and debugged production incidents across Lambda, Aurora, and DynamoDB.
- Owned CI/CD, infrastructure automation, and cloud operations for analytics products — the Terraform, AWS, and Python grounding the platform work is built on.
- Hardened deployment and certificate-renewal processes on legacy stacks and handled cross-team reliability and access management.
Highlights
Selected work
Self-service role vending
A PR-driven system for cross-account CI credentials: engineers describe the access they need in YAML, Terraform provisions short-lived OIDC roles, and no static keys ever leave the account. Turned a ticket queue into a merge.
Observability & incident flow
Datadog monitors for payroll Lambdas and SQS DLQs defined as code, plus a roadmap to consolidate paging into Datadog Incident Management — signals that page the right team instead of dashboards nobody reads.
SBOM governance programme
Software bill-of-materials generation and policy enforcement across build pipelines with JFrog Xray and CycloneDX, wired in as a custom pipeline step so provenance is checked before anything reaches production.
Compliance-as-code scanning
Checkov-based IaC scanning with findings mapped directly to SOC2 and CIS controls, so security posture is enforced at merge time instead of audited after the fact.
Cross-account event replication
An EventBridge architecture replicating events across AWS accounts with tested bidirectional loop-prevention — the connective tissue for a distributed, multi-account platform.
This page, as infrastructure
One self-contained static file — no build step, no framework. Designed to ship on S3 + CloudFront + ACM + Route 53, Terraform-provisioned, deployed by GitHub Actions via OIDC with no long-lived AWS keys. Practising what the day job preaches.
Skills
Tools of the trade
Cloud & compute
AWS, ECS Fargate, Lambda, Step Functions, EventBridge, SQS
Infrastructure as code
Terraform, GitOps, Docker, Kubernetes (CKA in progress)
Observability
Datadog, OpenTelemetry, monitor-as-code, incident management
CI/CD
Bitbucket Pipelines, GitHub Actions, OIDC-based delivery
Languages & scripting
Python, Bash, SQL
Security & compliance
SOC2, CIS, SBOM (JFrog Xray / CycloneDX), cross-account IAM, Lake Formation
Contact
Get in touch
Open to senior SRE, platform, and observability roles — UK relocation with sponsorship, or remote. Happy to walk through any of the work above.