// glossary
Plain-English definitions.
Operator-grade vocabulary for fractional CTO, AI leadership, MCP, agents, and the production stack around them. No jargon stack.
Agent Orchestration
Agent orchestration is the coordination layer that lets multiple AI agents share work, avoid collisions, and produce a single coherent outcome. It covers task dispatch, lane or queue management, handoffs, contracts between agents, and the operator authority that supervises all of it.
read the definition →AI CTO
An AI CTO is a chief technology officer at a company whose primary technical bets are AI-shaped. The accountabilities are the same as any CTO — architecture, hiring, vendor selection, roadmap, operations — but the surface area is dominated by models, evals, agents, and AI-specific risk.
read the definition →AI Pilot
An AI pilot is a small, scoped, measured experiment built to prove whether an AI bet clears a defined bar — accuracy, cost, latency, safety, operator judgment — before the team commits to the production investment that would land it permanently.
read the definition →AI Readiness Audit
An AI readiness audit is a focused assessment of a company's data, codebase, team, risk surface, and existing tooling against a specific AI bet — answering whether the company is actually ready to ship that bet, and what needs fixing before it can.
read the definition →AI Roadmap
An AI roadmap is a sequenced plan of specific AI bets the company will run over the next quarter or two — each with a measurable bar, an honest cost estimate, and a kill condition. A roadmap is decisions, not slideware: one pilot, three follow-on bets, and a clear order.
read the definition →AI Transformation
AI transformation is the multi-year program by which a company changes its products, operations, and culture around AI. It is bigger than a pilot, bigger than a roadmap, and not finished until the change shows up in how the company hires, ships, and operates.
read the definition →Chief AI Officer
A Chief AI Officer is an executive accountable for the company-wide AI thesis: strategy, risk, regulation, vendor selection, and the narrative the board hears about AI. The role usually reports to the CEO and exists at companies where AI is a regulated, multi-business-unit concern.
read the definition →Field CTO
A field CTO is a senior technical role inside a vendor whose job is to meet customers as a peer of their CTO — running strategic selling motions, architecting integrations, leading design partnerships, and feeding what they hear back into product. The role is internal to a vendor, not external to a customer.
read the definition →Fractional AI CTO
A fractional AI CTO is a senior technology leader who works with a company part-time, where the bets being made are AI-shaped — agents, large language models, multi-agent systems, retrieval, and the evaluation infrastructure around them — not just classical platform engineering.
read the definition →Fractional CTO
A fractional CTO is a senior technology leader who works with a company on a part-time, ongoing basis — usually one to three days a week — to set technical direction, design the architecture, and steer the engineering team without taking a full-time seat.
read the definition →Head of AI
A Head of AI is the functional lead accountable for the AI team, model quality, the evaluation harness, and the AI roadmap. The role usually sits inside engineering or data, reports to the CTO or VP, and owns the AI team's outputs the way a Head of Backend owns the backend.
read the definition →Interim CTO
An interim CTO is a senior technology leader who steps into a full-time CTO role for a fixed runway — usually three to nine months — covering a defined gap: a CTO departure, a migration with a hard deadline, or a fundraising window that demands technical leadership at the executive table.
read the definition →MCP Server
An MCP server is a process that exposes a set of tools, resources, and prompts to LLM hosts — like Claude, Cursor, or an agent runtime — over the Model Context Protocol. It is the standard wire between an AI agent and the systems it is allowed to act on.
read the definition →Model Context Protocol
Model Context Protocol — MCP — is an open standard for connecting LLM hosts to external tools, resources, and prompts. It defines the wire between an AI host and a server that exposes capabilities: discovery, schema, invocation, streaming, cancellation, and notifications.
read the definition →Multi-Agent System
A multi-agent system is a software system in which two or more AI agents act with their own scope and tools, coordinated to produce a shared outcome. Each agent owns a slice of the work — a role, a lane, a queue — and the orchestration layer keeps them from stepping on each other.
read the definition →Senior AI Systems Architect
A senior AI systems architect is the engineer responsible for the architecture of AI-shaped systems: model topology, retrieval and data layer, tool surface (often MCP-based), agent orchestration, evaluation, and the observability spine threaded through every layer.
read the definition →