# Naman Kumar

Lead Site Reliability Engineer · Melbourne, VIC, Australia

I lead reliability engineering for large AWS estates: production operations, observability, infrastructure automation and the teams that run them. Lately that includes practical AI tooling for incident response on AWS Bedrock.

## Summary

Lead Site Reliability Engineer with 7+ years of hands-on experience designing, scaling and automating large-scale distributed systems on AWS. Currently leading a team of 10 engineers supporting 80+ AWS accounts for a major financial services client. Proven track record in cutting incident response times, consolidating observability platforms, and moving ClickOps estates to Infrastructure as Code. Now focused on AI-driven operations, including LLM-powered incident tooling and secure enterprise adoption of Claude on AWS Bedrock.

## Key numbers

- **7+** years in production engineering: Cloud automation and SRE since Jan 2019, across product, agency and consulting roles.
- **80+** AWS accounts supported: Infrastructure support for a financial services client, delivered by the 10-engineer team I lead.
- **~880M** log events per week migrated: Splunk to Datadog at Cvent, unifying logs, metrics and traces on one platform.
- **50k+** concurrent users at launch: In the first 10 minutes of a Netflix game launch at AKQA, on Kubernetes clusters I designed.

## Experience

- **Lead Site Reliability Engineer**, Viable Solutions (via Synechron, for Latitude Financial Services), Aug 2025 – Present: Leads a 10-engineer team supporting 80+ AWS accounts, 24x7
- **DevOps Lead Consultant**, CI&T (for iSelect / CompareTheMarket), May 2025 – Jun 2025: CloudFormation to Terraform through the iSelect / CompareTheMarket merger
- **Lead Site Reliability Engineer**, AKQA, Sep 2024 – Apr 2025: SRE lead for a Netflix game launch: 50k+ concurrent users in 10 minutes
- **Senior Site Reliability Engineer**, Cvent, Oct 2019 – Sep 2024: Co-developed Sev1 automation (20+ min faster response); ~880M logs/week to Datadog
- **Cloud Automation Associate**, Infraguard, Jan 2019 – Oct 2019: Lambda automation and private-subnet patching across multi-cloud VMs

Full detail: [resume](https://naman-kumar2397.github.io/resume.md)

## Case studies

- [Alfred: a multi-agent AI SRE](https://naman-kumar2397.github.io/projects/alfred.md) (Latitude Financial Services (client, via Viable Solutions and Synechron)): Live in production on Amazon Bedrock AgentCore, orchestrating AWS DevOps Agent
- [It's What's Inside: The Game](https://naman-kumar2397.github.io/projects/netflix-game-launch.md) (AKQA · for Netflix): 50k+ concurrent users in the first 10 minutes
- [Large-scale observability migrations](https://naman-kumar2397.github.io/projects/observability-migrations.md) (Cvent (Splunk to Datadog) · Latitude (Sumo Logic to Datadog)): ~880M and 20M+ log events per week moved to Datadog
- **Claude for developers via a Bedrock proxy** (Latitude Financial Services (client, via Viable Solutions and Synechron)): GitHub Copilot on Claude, with guardrails and per-developer cost attribution

## Side projects

- **Buildkite Build Watcher**: Open-source Chrome extension (Manifest V3), live on the Chrome Web Store. Plays a distinct synthesised chime when a build passes, fails or is blocked on input, and auto-watches builds you trigger. ([Chrome Web Store listing](https://chromewebstore.google.com/detail/buildkite-build-watcher/ppaolcakoopkiaeigaeamcakhafiicii), [Source on GitHub](https://github.com/naman-kumar2397/chrome_buildkite))

## Capabilities

- **Site Reliability Engineering**: SLI/SLO standards, zero-downtime rollouts, Sev1 automation that cut response time by 20+ minutes
- **Observability**: Two large log migrations to Datadog (~880M and 20M+ events per week); GPU monitoring for a launch
- **Cloud Architecture on AWS**: 80+ AWS accounts supported; 70+ ECS clusters; Rackspace to AWS migration
- **Infrastructure as Code and Platform**: ClickOps to IaC, CloudFormation to Terraform, Octopus Deploy HA at 99.9%
- **Technical Leadership**: Leads 10 engineers with a follow-the-sun 24x7 model; mentored 4 engineers
- **AI-assisted Operations**: Alfred, a multi-agent AI SRE live on Bedrock AgentCore; a governed Claude proxy on Bedrock

## Contact

- Email: naman.kumar2397@gmail.com
- LinkedIn: https://www.linkedin.com/in/-namankumar/
- Website: https://naman-kumar2397.github.io/
- Resume (PDF): https://naman-kumar2397.github.io/resume.pdf