Twofold builds systems that build systems. Two humans, hundreds of agents, thirty years of enterprise engineering — shipping production software at a pace that shouldn't be possible for a team this size. If it's a system, we can automate it. Everything on this site proves it.
Five products public. More on the bench.
A Claude-native team of AI specialists for every knowledge worker.
M365-wired, governance-aware, ledger-backed AI specialists. Built on Claude Code primitives, for firms that need AI to produce output, not prompts.
Cloud ops for the team without a platform engineer.
AI-native observability, security, cost, IaC drift, and runbook automation for Azure. Detect, explain, and act — priced for the middle market.
Plain language in. Production-grade software out.
Describe a workflow in plain language. A crew of Claude agents designs, builds, and load-tests it in your workspace. Nothing ships until you set the keystone.
The institutional memory you own, not rent.
A sovereign learning loop: knowledge as plain-text OKF atoms in your own git, served over GraphRAG and MCP, curated by people, extended by agents.
A self-replicating plugin marketplace for Claude Code.
A Claude Code plugin marketplace built as a software factory. 13 MIT-licensed plugins live, every one produced and governed by the pipeline itself.
Build notes, research, and strong opinions from the people shipping the products. The agents don't get bylines. Yet.
Three flavors of drift, what terraform plan actually catches, the free tools that fill the gaps, and the point at which the homemade approach breaks down.
READ POSTMost Azure subscriptions run VMs at p95 CPU under 40%. The deterministic ruleset for right-sizing them, the script that finds them, and the commit math.
READ POSTWhy a two-person AI lab built its own cloud ops platform for Azure: detect, explain, act — AI-native by design, priced for teams without a platform engineer.
READ POSTDakota and Dillon Smith — thirty years of enterprise systems between them — plus the agent teams they build to do the work of a company many times this size.
Systems architect. AI/agentic architecture and the structural engineering that makes software production-ready.
Engineering leader. Cloud infrastructure, platform engineering, and the operational rigor that ships systems.
Everything on this site was shipped by two humans and hundreds of agents. That same system takes outside work: agentic systems, workflow automation, production software. Scoped to an outcome, shipped at the pace the portfolio proves. No 200-slide strategy decks. No six-month discovery. We build.