Stay a While and Listen
I'm Deckard Blaine. I build AI deployment kits for back-office and professional services teams: finance, operations, legal, HR, and beyond.
These are deployment kits - complete, structured systems that turn Claude into a genuine working partner for the things you actually do all day: variance analysis, board reporting, meeting documentation, forecasting, file management, internal communications, and more. Each one is designed for a specific professional function, not a generic use case.
Each kit includes a tuned system prompt, role-specific templates where the workflow demands them, and a deployment guide. You configure once, then work with it like a colleague who is fast, thorough, and never too busy to help.
A note on what these tools are and are not: AI models can produce output that is confident, fluent, and wrong. That is a known property of the technology. These kits are designed to make the output better, more consistent, and more useful - but they do not eliminate the need for you to review what comes back.
Why This Exists
I have spent my career in finance, in environments where the numbers are never just numbers. Budgets are narratives. Forecasts are positions. Board decks are statecraft.
I started building these tools for myself because the gap between what AI could theoretically do and what it actually did for a working professional was embarrassing. The generic tools did not understand my domain. The prompting guides assumed I had time to become a prompt engineer. I did not.
The kits closed that gap. So I started documenting them properly, and now I'm sharing them.
What You Get
Every kit follows the same architecture I use in my own workflow and deploy across the teams I work with. It is adding value to real work in a real business, and that is why I am sharing it. They are designed for Claude on the Pro, Max, Team, or Enterprise plan - no API, no terminal, no code required. If you can paste text into a chat window, you can deploy these.
The Mechanical Clerk is always free, no membership required - a complete meeting documentation kit with system prompt, template library, and deployment guide. Deploy it before your next meeting and see what a purpose-built kit actually does.
Analyst subscribers also receive a monthly operational essay on deployment methodology and design thinking. Strategist subscribers receive behind-the-build breakdowns examining how specific kits were designed and why.
Membership tiers range from The Foundation (7-day free trial) to The Quorum (capped at 25 founding members). Full tier details, pricing, and benefits are on the main page.
The library launched in May 2026 with three kits and grows monthly. Current and planned coverage spans finance and accounting, legal and compliance, operations, HR and people, sales and revenue, marketing and communications, strategy and leadership, analytics and reporting, customer success, security and privacy, knowledge management, and AI operations - with room to expand as the work demands it.
The Design Philosophy
In 1979, IBM produced a training slide that said: "A computer can never be held accountable, therefore a computer must never make a management decision." That was written about mainframes. It is more true now. Two convictions follow from it, and they govern everything here:
The kits amplify your judgment. They never substitute for it. You remain the accountable decision-maker - the one who reviews the output, applies the context no system has, and signs for the result.
Each kit works on its own. If you choose to deploy more than one, you become the connective tissue between them - directing each, carrying context from one to the next, and deciding what matters. That is not a limitation of the system; it is the system working as designed.
In my experience, in a back-office environment - such as finance, legal, HR, operations - the human-as-orchestrator is the stronger architecture. Others will disagree; there is an entire field of research devoted to having AI agents manage other AI agents, and in software engineering, where most AI platforms today are purpose-built, that argument is well developed. But professional environments are not engineering departments. The work is less structured, the stakes are more ambiguous, and the judgment calls depend on context that no system has. These kits are built on the conviction that the professional in the chair is the right person to direct the work, review the output, and make the call. The work the kits produce is the argument for that position.
This also means the technical barrier has been carefully designed to be as low as possible. You do not need to learn a new platform, adopt a developer workflow, or acquire skills outside your domain to use what is here. If the tools required expertise their operators did not already have, they would fail the people they were built for.
Who This Is For
Executives, department heads, and the professionals who report to them - in finance, operations, legal, HR, and any function where the work matters more than the tooling. If AI should be helping with your work but mostly isn't, it is probably because the tools were not built by someone who does that work.
These were.