Two components, one experience
MyCorum.ai has two distinct components that work together. Le Corum is the deliberation engine — five independent minds that analyze, confront, and converge. You observe Le Corum through the sidebar: read-only, unfiltered, in real time.
MyPilot is everything else. It is the conversational interface that sits between you and Le Corum. It speaks in plain language. It asks clarifying questions. It recommends the right level of analysis. It narrates what Le Corum is doing as it happens. And when Le Corum delivers its verdict, MyPilot walks you through it.
Without MyPilot, Le Corum would be a powerful engine with no cockpit. MyPilot is the cockpit.
Before the deliberation: MyPilot frames
The quality of a deliberation depends on the quality of the question. A vague question produces a vague verdict. Before launching Le Corum, MyPilot helps you identify what you are actually deciding — not just what you typed.
"Should we expand to the US market?" is a statement that contains multiple possible decisions: a timing question, a resource question, a strategic priority question, a competitive positioning question. Each requires a different analytical emphasis. MyPilot surfaces this before the deliberation begins, so Le Corum's five minds are analyzing the right problem.
MyPilot recommends — but you decide
Once the question is framed, MyPilot recommends the deliberation mode that fits your need. This recommendation is built on a principle that distinguishes MyCorum.ai from most SaaS products: MyPilot will never recommend a more expensive mode than your question requires.
If your question can be fully answered by an Express routing for 1.2 credits, MyPilot says so. If it recommends The A-Team at 3 credits, it explains why Express would miss something important. If you override and ask for The Dream Team — the most intensive deliberation, up to 6 adaptive rounds — MyPilot executes without question.
You know your decision better than any AI system. MyPilot's role is to give you accurate information about the tradeoffs, then get out of your way.
During the deliberation: MyPilot narrates
While Le Corum deliberates, you observe its progress through the sidebar. But Le Corum operates in technical rounds — Independent analysis, Confrontation, Resistance test, Assumption audit. Without translation, this is opaque.
MyPilot narrates each round in plain language as it completes: what happened, what was found, what disagreements emerged, what the next round will address. If the Biodiversity Narrator detects dangerous convergence and triggers an additional challenge round, MyPilot tells you why. If a HITL pause is activated — Le Corum needs your input to continue — MyPilot alerts you immediately and tells you exactly what it needs.
You are never in the dark about what is happening inside the deliberation. That transparency is structural, not cosmetic.
After the deliberation: MyPilot interprets
When Le Corum delivers its Corum Synthesis, the output is rich and structured — but dense. A GO recommendation with confidence 8.2/10, a decision matrix with four dimensions, three next steps with owners and deadlines, an information gap, and a minority report from The Strategist. The information is all there. Making sense of it quickly is the challenge.
MyPilot walks you through it. It highlights the confidence score and what it means in practical terms. It flags the minority report and explains why you should read it before deciding. It summarizes the key conditions that could invalidate the recommendation. And it reminds you, at the end, that the decision is yours.
The 16 agents behind MyPilot
MyPilot is not a single AI assistant. It is an orchestrated layer of 16 specialized agents operating across four stages of the deliberation lifecycle.
In the pre-deliberation layer, a Decision Framing Coach identifies the real decision beneath the stated question. A Mode Advisor recommends the right deliberation intensity. An Empathy Gate recognizes when the decision carries emotional weight and adjusts the conversation accordingly.
During deliberation, a Context Coach manages HITL pauses and requests additional information from you when Le Corum needs it. A Round Narrator translates each completed round into plain language. A Tension Narrator flags deep disagreements when they emerge. A Biodiversity Narrator alerts you when dangerous consensus is detected.
After deliberation, a Synthesis Herald delivers the verdict summary. A Confidence Interpreter explains what the score means in practical terms. An Outcome Prompter follows up 30 days later to learn what happened — feeding that outcome back into Le Corum's calibration for future deliberations.
None of this is visible as complexity. MyPilot feels like one coherent voice. That is the point.