Reference
Glossary
What we mean when we say it — one definition, one right usage, one wrong one. Brainis uses the same names on the site and inside the product; if a term shows up in the app, it's defined here.
- Brainis
- The product, company, system.
- Right: Brainis compiles your strategy into missions.Not: the brainis platform enables…
- Cortex
- The supervisory intelligence: reads state, asks the deciding question, models options, compiles missions, supervises, escalates.
- Right: Cortex flags the binding constraint in your Monday brief.Not: Chat with Cortex AI, your virtual assistant!
- Company State
- The live, typed, sourced model of the business.
- Right: Every claim in your Company State carries a source and a freshness date.Not: We store all your data in one place.
- Objective
- A measurable desired change with a horizon. Retires “goal” in product copy.
- Right: Objective: 10 design partners in 30 days.Not: Set your goals and dreams!
- Strategy
- The chosen theory of how to reach an objective; versioned; approval creates a Decision.
- Right: Approve the strategy and Brainis compiles it into missions.Not: Download your strategy PDF.
- Strategy Engine
- The survey-driven system (A-to-Z or Together modes) producing the eight connected plan artifacts. Same product formerly “Strategy OS.”
- Right: The Strategy Engine asked three questions that changed our pricing.Not: our AI business plan generator
- Blueprint
- The connected artifact set projected from a strategy (business plan, brand, marketing, media, website, agent tools, GTM, lead strategy).
- Right: Edit the target segment; the blueprint shows six sections that must change.Not: your documents folder
- Mission
- The outcome-bound executable unit: outcome, owners, budget, acceptance criteria, verification, measurement.
- Right: The launch mission is waiting on one pricing decision.Not: using “mission” for the gamified rewards hub (that renames to Progress)
- Mission Control
- The delivery orchestration layer (formerly “Delivery OS” — retired as bad-sounding by founder).
- Right: Mission Control routed the research to a reasoning agent and the build to a coding agent.Not: our delivery OS pipeline
- Play
- A reusable strategy/execution template (the 439+).
- Right: Run the ‘revive stalled deals’ play.Not: automation (plays can contain human decisions and verification gates)
- Agent
- An AI worker (any provider) acting under a contract. Never Brainis itself; never a Brainis employee.
- Right: A coding agent fixed the signup bug; a different verifier re-ran the tests.Not: our AI employees
- Agent Fleet
- The provider-neutral runtime (public name; “Brainis Runtime” on technical pages).
- Right: The Agent Fleet runs Claude, GPT, and Codex as workers — Brainis stays the employer.Not: our multi-model router
- Brainis MCP
- The governed context/tool layer; the only way agents touch company systems.
- Right: The agent's MCP grant allowed reading interviews and writing CRM drafts — nothing else.Not: agents have full access to your workspace
- Verify
- The independent quality layer (formerly “Quality OS”); produces the “Verified by Brainis” mark.
- Right: Verify rejected the landing page for an unsupported claim.Not: AI double-checks itself
- Authority Contract
- Scoped permission to act (actor, verbs, budget, risk, reversibility, window, expiry).
- Right: Outreach runs at L4 under a $200 weekly contract.Not: autonomy switch
- Autonomy Levels (L0–L7)
- Observe · Recommend · Prepare · Ask to act · Act in bounds · Run missions · Manage outcomes · Earned autonomy.
- Right: Pricing stays at L2; research runs at L5.Not: fully autonomous mode
- Outcome
- The measured state change vs. the prediction.
- Right: Predicted 8–14 qualified conversations; actual: 11.Not: calling a deliverable an outcome
- Decision Ledger
- The immutable state→recommendation→decision→outcome record.
- Right: The ledger shows what we believed when we chose this.Not: audit log
- Business World Model
- The proprietary predictive model (with Company State Encoder, Decision Policy Model; benchmarked by BusinessBench).
- Right: The World Model's prediction carries a range and a version.Not: our AI knows business better than anyone
- The AI Suite / [Domain] AI
- The department products: Hiring AI, People AI, Revenue AI, Finance AI, Work AI (work management only), Marketing AI, Product AI, Customer AI, Operations AI, Knowledge AI, Analytics AI (dashboards/reports/BI), Communication AI, Wellness AI.
- Right: Revenue AI is a real CRM — agents work inside it through MCP.Not: “Delivery AI” (rejected invented noun); “Work AI runs the OS” (dead positioning)
- Your team
- The customer's own humans inside missions.
- Right: Your team reviews the draft before it sends.Not: never “our team” in a delivery sentence
- Design Partner
- Early customer with roadmap influence, running Brainis themselves.
- Right: Design Partners get a direct line to the founder.Not: founding partner implementation package
- Progress
- The renamed gamified hub (challenges, badges, streaks). Frees “Mission” for real work.
- Right: Your Progress streak is at 12 days.Not: calling the gamified hub a Mission