Technology stack

Shipped at real scale, with something you can go and read. Every entry here links to the case study or to this site's own architecture.

1. 01 Kubernetes Platform & Infrastructure Production clusters at enterprise scale; the load-test rig drove one to 100GB/s. the receipt →
2. 02 Cloudflare Workers Platform & Infrastructure One Worker serves this entire site — SSR, every API route, and the MCP server. the receipt →
3. 03 Workers AI Platform & Infrastructure Edge inference behind the Copilot, routed through AI Gateway. the receipt →
4. 04 TypeScript Platform & Infrastructure End-to-end typed, astro check clean on every commit. the receipt →
5. 13 Terraform Delivery & Automation Multi-cloud IaC; carried the DataDog-to-ELK migration. the receipt →
6. 14 Ansible Delivery & Automation Configuration management alongside Terraform on the observability build. the receipt →
7. 15 GitHub Actions Delivery & Automation Ran the automated content pipeline end to end. the receipt →
8. 18 Prometheus Data & Observability Metrics backbone for the infrastructure-recommendation agent. the receipt →
9. 19 ELK Data & Observability Self-hosted, ingesting 5TB of logs a day after leaving usage-priced vendor billing. the receipt →

Run in production, without a public write-up to point at.

1. 05 AWS Platform & Infrastructure Multi-account estates, migrations, and the usual cost-control work.
2. 06 Azure Platform & Infrastructure Enterprise workloads, mostly alongside on-prem.
3. 07 GCP Platform & Infrastructure Data and ML-adjacent workloads.
4. 08 K3s Platform & Infrastructure Lightweight clusters where full Kubernetes is overhead.
5. 09 Docker Platform & Infrastructure Still the right answer more often than a cluster is.
6. 10 Go Platform & Infrastructure Services and tooling where the runtime matters.
7. 11 Python Platform & Infrastructure Automation, data work, and most AI plumbing.
8. 16 GitLab CI Delivery & Automation Gates this site: typecheck, lint, tests, dry-run, Lighthouse, secret scan.
9. 20 Grafana Data & Observability Dashboards over Prometheus and Loki.
10. 21 PostgreSQL Data & Observability Default relational choice; hybrid BM25 + pgvector search without a vector DB.
11. 22 Redis Data & Observability Caching and queues where latency is the constraint.
12. 25 Vault Security Secret management and dynamic credentials.

Built something real with it. Not yet a default choice.

1. 12 Istio Platform & Infrastructure Service mesh rollout; powerful, and rarely worth its operational cost.
2. 17 Pulumi Delivery & Automation Real programming languages for infra; the tradeoff is who can review it.
3. 23 LangChain Data & Observability Fine for a prototype; production agents wanted typed tools and explicit control flow.
4. 24 Kafka Data & Observability Right at real volume, heavy well before it.
5. 26 OPA Security Policy as code — good when policy is genuinely shared across teams.
6. 27 Falco Security Runtime detection in clusters.
7. 28 Trivy Security Image and dependency scanning in CI.

Know it, have run it, would not start here today.

// nothing here yet — an invented entry would be worse than an empty ring