AI Assistants (ChatGPT, Claude, Replit)Various
AI tooling applied to the specific operational tasks where it improves speed and accuracy, directed by senior judgment.
AI is part of the operational discipline now. The honest framing is that AI is an accelerant for the parts of the work where it actually accelerates: research synthesis, code generation, copy iteration, document analysis. It is not a substitute for senior judgment about what to build, when to build it, or whether the underlying strategy is correct. We are deliberate about where AI lives in the work.
Operator prompts informed by domain expertise, plus context from the specific engagement: documents, code, data, and the strategic frame the operator is working inside.
Drafted code, drafted copy, structured analysis, and the iterative work that an experienced operator then refines and decides whether to ship.
Problems it solves
- >Code drafting for integration logic, Apps Script automation, and internal tooling.
- >Research synthesis across documentation, transcripts, and prior engagement notes.
- >Copy iteration where the operator brings the strategic frame and AI accelerates the production.
- >Document analysis at scale for audit and discovery work.
Where it earns the line item.
Used inside a discipline, AI compounds operator productivity. Used outside a discipline, AI produces confident-sounding output that misses the architectural questions that actually matter. We use it the first way and have written publicly about why the second way damages the work.
What we have run into in real engagements.
- !Confident hallucination is the failure mode that catches teams without domain expertise to verify output.
- !Strategic decisions are not AI-shaped problems. Tools that pretend otherwise produce expensive misdirection.
- !Vendor and model selection matters; the right tool for the job varies by use case.
- !Privacy and data handling discipline is non-negotiable, especially for client work.
Where AI Assistants (ChatGPT, Claude, Replit) fits in Define, Develop, Deliver.
AI-assisted research, framework analysis, and the document discovery that informs senior decisions.
Code drafting, copy iteration, automation scripting under operator review.
Reporting analysis, dashboard prose, and the documentation work AI accelerates.
Other platforms that show up in the same conversations.
Related services and operations work
Frequently asked questions
QAre you AI-skeptical or AI-enthusiastic?+
QDo you work in AI Assistants for clients who already use it?+
QCan you help us decide whether AI Assistants is the right tool for us?+
Need senior help with AI Assistants (ChatGPT, Claude, Replit)?
Most AI Assistants (ChatGPT, Claude, Replit) engagements we run are inside instances that have been operating for a while and have accumulated configuration drift. The work is to repair, re-govern, and make the platform behave the way the strategy assumes it does.