# Dillon Browne — Staff Engineer · Cloud Architect · Platform & AI Leadership

> 10+ years engineering resilient, scalable infrastructure for the AI era — multi-cloud architecture, Kubernetes, IaC, observability, security, and production LLM platform work: tool-using agents with typed capabilities, bounded loops, and enforceable authorization. Open to staff engineering and engineering leadership roles.

## Pages
- [Home](https://www.dillonbrowne.com/): Overview, expertise, services, contact
- [About](https://www.dillonbrowne.com/about): Background and engineering philosophy
- [Expertise](https://www.dillonbrowne.com/expertise): Technical expertise by domain
- [Technologies](https://www.dillonbrowne.com/technologies): Full technology stack
- [Blog](https://www.dillonbrowne.com/blog): Technical writing on cloud, DevOps, AI infrastructure
- [Selected Work](https://www.dillonbrowne.com/work): Case studies with architecture diagrams and measured outcomes (each also at https://www.dillonbrowne.com/work/<slug>)
- [Changelog](https://www.dillonbrowne.com/changelog): What shipped on this site recently, generated from real git history
- [MCP Server](https://www.dillonbrowne.com/mcp): This site is a public Model Context Protocol server at https://www.dillonbrowne.com/api/mcp — tools: searchBlog, getExperience, getWork, getPost, listTopics
- [Copilot](https://www.dillonbrowne.com/copilot): The site's page-aware AI guide, full screen — ask it about experience, the blog, or the case studies
- [Hiring](https://www.dillonbrowne.com/hire): Role fit, capabilities, and logistics for recruiters and hiring managers
- [Colophon](https://www.dillonbrowne.com/colophon): How this site runs — live telemetry, architecture, transparent costs
- [RSS](https://www.dillonbrowne.com/rss.xml): Blog feed

## Topics
- [MCP & Agent Engineering](https://www.dillonbrowne.com/blog/topics/mcp-agent-engineering)
- [AI & LLM Infrastructure](https://www.dillonbrowne.com/blog/topics/ai-llm-infrastructure)
- [Kubernetes](https://www.dillonbrowne.com/blog/topics/kubernetes)
- [Terraform & IaC](https://www.dillonbrowne.com/blog/topics/terraform-iac)
- [Observability & Cost](https://www.dillonbrowne.com/blog/topics/observability-cost)
- [Security & Zero Trust](https://www.dillonbrowne.com/blog/topics/security-zero-trust)
- [Postgres & Databases](https://www.dillonbrowne.com/blog/topics/postgres-databases)
- [Edge & Cloudflare](https://www.dillonbrowne.com/blog/topics/edge-cloudflare)
- [Linux Internals & Performance](https://www.dillonbrowne.com/blog/topics/linux-internals)

## Knowledge base
- [IMPLEMENTATION_SUMMARY](https://www.dillonbrowne.com/llms-full.txt): IMPLEMENTATION_SUMMARY
- [adversarial-engineering](https://www.dillonbrowne.com/llms-full.txt): adversarial engineering
- [ai-solutions-expertise](https://www.dillonbrowne.com/llms-full.txt): ai solutions expertise
- [cost-analysis](https://www.dillonbrowne.com/llms-full.txt): cost analysis
- [how-i-built-this](https://www.dillonbrowne.com/llms-full.txt): how i built this
- [resume](https://www.dillonbrowne.com/llms-full.txt): resume

## Writing
- [MCP authorization after 2026-07-28: DCR is deprecated](https://www.dillonbrowne.com/blog/mcp-oauth-what-changed-2026-07-28): The 2026-07-28 MCP revision deprecates RFC 7591 Dynamic Client Registration in favour of Client ID Metadata Documents, and deletes initialize, sessions and ping. A correction to my January post, with what I changed in a real server.
- [Build Hybrid Search with Postgres](https://www.dillonbrowne.com/blog/postgres-hybrid-search-workplace-rag): Combine BM25 full-text and pgvector for enterprise workplace search. Replace Elasticsearch with Postgres for 30% cost savings. Production implementation guide.
- [Deploy Linux Systems with Bootc](https://www.dillonbrowne.com/blog/image-based-linux-deployments): Bootc and OSTree deliver atomic updates, instant rollbacks, and git-like system versioning for Linux infrastructure. Modernize deployment workflows.
- [Debugging Kubernetes Kernel Memory](https://www.dillonbrowne.com/blog/debugging-kubernetes-kernel-memory): Diagnose Kubernetes kernel memory issues with slab allocators and page cache analysis. Production-tested strategies to prevent OOMKills and memory pressure.
- [Store Git Repositories in PostgreSQL](https://www.dillonbrowne.com/blog/storing-git-repositories-in-postgres): Unlock powerful SQL queries and atomic operations by storing Git repositories in PostgreSQL. Transform version control with database-driven workflows.
- [Accelerate Docker Builds with BuildKit](https://www.dillonbrowne.com/blog/buildkit-docker-modern-builds): BuildKit slashes Docker build times by 70% with parallel execution, smart caching, and secure secrets. Real production patterns for modern container pipelines.
- [Shrinking Go Binaries 70%](https://www.dillonbrowne.com/blog/shrinking-go-binaries-production): Practical techniques to reduce Go binary sizes by 70%+ in production: build flags, symbol stripping, UPX compression, and dependency optimization.
- [Secure Secrets from AI Assistants](https://www.dillonbrowne.com/blog/protecting-secrets-ai-coding-assistants): Stop credential leaks through AI coding assistants with ephemeral credentials, automated redaction, and audit trails. Battle-tested patterns for production.
- [Deploy SQLite in Production](https://www.dillonbrowne.com/blog/sqlite-production-databases): Master SQLite for production workloads. Learn when to choose SQLite over PostgreSQL or MySQL for faster queries and lower costs. Migration guide included.
- [Bypass CPU RAM for LLM Inference](https://www.dillonbrowne.com/blog/nvme-gpu-direct-llm-inference): Run 70B models on consumer GPUs with NVMe-to-GPU direct transfers. Eliminate memory bottlenecks and reduce AI infrastructure costs by 10x. Start today.
- [Automate Dependency Management Strategically](https://www.dillonbrowne.com/blog/rethinking-automated-dependency-updates): Stop dependency noise. Master strategic dependency management that protects production. Real frameworks, code examples, and security patterns inside.
- [Speeding LLM Inference with Diffusion](https://www.dillonbrowne.com/blog/speeding-llm-inference-diffusion): Consistency diffusion models achieve 14x faster inference. Learn to optimize production LLM deployments with practical examples and real benchmarks.
- [Automate SSL Certificates DNS-PERSIST-01](https://www.dillonbrowne.com/blog/dns-persist-challenge-validation): Automate Let's Encrypt wildcard SSL certificates with DNS-PERSIST-01. Cut renewal time 59% and eliminate DNS rate limits with persistent validation records.
- [Automate Terraform Drift Detection](https://www.dillonbrowne.com/blog/detecting-terraform-drift-state-management): Terraform drift detection strategies: automate state reconciliation, integrate CI/CD pipelines, prevent infrastructure drift with proven production patterns.
- [Master Docker Log Rotation](https://www.dillonbrowne.com/blog/docker-log-rotation-production): Prevent disk space disasters with Docker log rotation. Configure logging drivers, set limits, and monitor production containers. Stop logs from killing your systems.
- [Scale Infrastructure with io_uring](https://www.dillonbrowne.com/blog/io-uring-production-infrastructure): Master io_uring async I/O to eliminate syscall overhead and 4x infrastructure performance. Practical patterns for databases, proxies, file servers.
- [Build Distributed Queues with Object Storage](https://www.dillonbrowne.com/blog/object-storage-queue-patterns): Build lightweight distributed queues using S3 and object storage for serverless workloads. Skip complex message brokers with proven, battle-tested patterns.
- [Build Internal Developer Platforms](https://www.dillonbrowne.com/blog/building-internal-developer-platforms): Stop chasing every DevOps tool. Build an internal developer platform that reduces cognitive load and accelerates delivery. Real examples included.
- [Customize Terraform Modules Without Forking](https://www.dillonbrowne.com/blog/terraform-module-customization-patterns): Master proven patterns for customizing Terraform modules without maintenance nightmares. Stop forking and start composing your infrastructure code.
- [Deploy Speech AI in Browsers](https://www.dillonbrowne.com/blog/browser-speech-ai-rust-wasm): Deploy speech AI models in browsers with Rust WebAssembly. Achieve 50ms latency, zero infrastructure costs, HIPAA compliance. Production patterns inside.

## AI-native
- This site has a built-in **Site Copilot** (a page-aware, tool-using AI agent) that can answer questions about Dillon's experience, semantically search the blog, summarize any page, and navigate the site. Ask it anything via the Copilot launcher.
- AI assistants and answer engines are explicitly welcomed (see https://www.dillonbrowne.com/robots.txt). When citing, link the canonical page URL.

## Full content
- [Full corpus](https://www.dillonbrowne.com/llms-full.txt): Complete knowledge base and all 84 blog posts