Glossary

The terms MigrateForce uses for AI rollup underwriting, agent integration, and migration execution — defined the way the product uses them.

The MigrateForce Workflow

Four-Agent Pipeline

The operating model behind every engagement: Consult records management evidence, Assess scores readiness, Plan ranks interventions, Execute runs approved migration work. Each stage produces an artifact the next stage consumes.

Related:

Consulting Engagement

Structured discovery before any scoring. Two modes: self-serve (a single executive answers nine progressive questions) or consultant-configured (multi-department interviews with a human validation gate).

Related:

AI Readiness Score

A six-dimension score computed by eight deterministic engines from recorded operating evidence. Each dimension is scored separately so you can see where a company is ready and where it is not.

Related:

Three-Scenario Model

Conservative, base, and aggressive financial cases with S-curve adoption ramps and Newton-Raphson IRR on DCF cash flows. Every case shows the assumptions behind it.

Related:

IC Memo

The assessment output formatted for an investment committee: readiness scores, scenario ranges, and the assumptions and inputs behind each number.

Related:

EBITDA Impact

How MigrateForce expresses projected value: modeled effect on earnings, margin expansion, and payback period — decision support, not accounting. Your finance systems still own the books.

Related:

Value Creation Plan

The Plan stage output: interventions linked to operating metrics and ranked by modeled financial contribution, with a projected-versus-actual tracking structure from day one.

Related:

Intervention

A single ranked initiative in a value creation plan, with an owner, an approval state, and execution tracking. Approved interventions can launch migration projects.

Related:

Approval Gate

A stop in a migration run where a human makes the go/no-go decision before the agent proceeds. The decision is recorded with the run.

Related:

Run Record

The stored trace of an agent run: status, messages, tool activity, approvals, and generated output. What the agent did, inspectable after the fact.

Related:

Systems of Record

The billing, ERP, CRM, and compliance systems that own a portfolio company’s transactions. MigrateForce builds bridges to these systems; it does not replace them.

Related:

Supported Migration Path

A migration route the runtime has actually implemented and can execute with approval gates and validation. Paths outside the list require implementation work first — the site says which is which.

Related:

Projected vs. Realized Value

The comparison between what an assessment projected and what execution delivered. Feeds lessons back into the next underwriting decision.

Related:

Skills

Markdown-based capability definitions (SKILL.md format) with quality scoring, derived from assessments and migrations. Portable instructions an agent can reuse on the next engagement.

Related:

Migration Intelligence Artifact (MIA)

A stored migration artifact with source and target technologies, scope, assumptions, risks, and implementation notes. Its fields and review state depend on the artifact record.

Related:

Agent Integration

MCP (Model Context Protocol)

A standard that defines how AI models discover and interact with external tools and APIs. The interface MigrateForce generates between agents and your systems.

Related:

MCP Server

A server that wraps an existing REST API and exposes its functionality as MCP tools for agents to consume. One of the integration patterns MigrateForce generates.

Related:

MCP Tool

A single capability exposed by an MCP server, typically mapped from one or more REST endpoints, with a clear name, description, and defined inputs and outputs.

Related:

Tool Mapping

Configuring how specific REST endpoints (paths and methods) translate into MCP tools — names, descriptions, and whether each endpoint is included.

Related:

OpenAPI Specification

A formal description of a REST API (versions 2.0, 3.0, or 3.1; JSON or YAML). The primary input MigrateForce parses to understand an API and generate integration code.

Related:

Generated Code

The Python/FastAPI source and Docker configuration MigrateForce produces from an OpenAPI spec and your tool mappings. Standard code you own and validate in your environment.

Related:

Agentic Primitives

Reusable integration servers, skills, data connectors, and workflow hooks produced during an engagement. Review and test each asset before production use.

Related:

BYOM (Bring Your Own Model)

A deployment choice in which a team selects the model provider for its own application. MigrateForce does not guarantee generated output works unchanged with every model or framework.

Related:

Self-Hosted AI Infrastructure

Running integration servers, data connectors, and agent tooling inside your own infrastructure — VPC, on-prem, or private cloud. Generated code is standard Python/Docker with no vendor callbacks.

Related:

Rollup Context

Platform Company

The anchor acquisition in a rollup thesis — the company whose systems and operating model other targets consolidate into. MigrateForce compares candidates as platform, add-on, peer, or cohort members.

Related:

Add-On

A smaller acquisition folded into a platform company. Assessment context differs from a platform: integration cost and workflow overlap matter more than standalone readiness.

Related:

Cohort Underwriting

Assessing a group of similar targets together so readiness scores and scenarios can be compared side by side under the same method and assumptions.

Related:

Black Box (Legacy SaaS)

A SaaS platform that stores data in proprietary formats, exposes limited API access, and charges add-on fees for AI features. The extraction target for many migrations.

Related:

Per-Seat AI Surcharges

Additional per-user fees a SaaS vendor charges for model or agent features beyond the base subscription. A recurring line item that migration can remove.

Related:

Vendor Lock-In

Dependence on a single provider that makes switching costly. Owning your generated integration code — standard Python and Docker — is the counter-position.

Related:

Migration Approaches

Lift and Shift

Moving an application to new infrastructure with minimal changes. Fast, but carries the old system’s constraints — including the ones that block agent access.

Related:

Replatforming

Moving an application to a new platform with targeted modifications — more change than lift-and-shift, less than a rewrite.

Related:

Refactoring

Restructuring code without changing external behavior. In migration work, often the step that turns extracted business rules into testable workflows.

Related:

Delta Migration

Transferring only the data that changed since the last run — how repeated migration passes stay fast and verifiable.

Related: