MiroFish

MiroFish Live

Run live AI prediction simulations from the material you already have.

MiroFish Live is for current decisions: upload or paste focused seed material, describe the scenario, generate a simulation graph, and review a report you can question like a research analyst.

Best first run: one brief, one decision question, one time horizon. MiroFish works better when the live scenario is narrow enough for the agents, graph, and final report to stay grounded.

What live use is good for

  • Testing how different groups may react to a launch, policy, market event, or public narrative.
  • Turning fresh notes, articles, transcripts, or internal briefs into structured scenario inputs.
  • Comparing optimistic, neutral, and stress-case branches before a decision is made.
  • Producing a report that can be questioned, refined, and reused for follow-up runs.

How a live scenario moves through MiroFish

1. Seed the world

Add the source material that defines the situation. A short, reliable brief beats a large mixed archive for the first live run.

2. Build the graph

MiroFish extracts entities, relationships, assumptions, and tensions so the simulation has a structured world to operate inside.

3. Run and question

Run the simulation, inspect the report, then ask follow-up questions about weak signals, risks, likely reactions, and alternate paths.

Live vs demo vs offline

MiroFish Live is the hosted route for practical runs. MiroFish Demo is the safest way to understand the interface and sample flow. MiroFish-Offline is for teams evaluating self-hosting, private deployment, or local model dependencies.

Related MiroFish guides