Prophet

AI forecast rooms for hard decisions

Stress-test the decisionbefore you make it.

Upload source material. Prophet runs it through structured debate and returns a forecast where disagreement stays visible and every claim links back to evidence.

Upload the source packet
Keep disagreement visible
Interrogate the brief in chat

Forecast room preview

Will energy prices spike if subsidy rules shift in Q3?

Room live

Forecast brief

Probability of a price spike within the next quarter

0%

medium confidence
Supporting0%
Skeptical0%
Unresolved0%

3 sources · 32 entities · 12 signal clusters

Live debate map

See where the room agrees, where it hesitates, and what still needs evidence.

Ask follow-up questions in chat

Main driver

Policy timing now dominates pricing assumptions

Most contested

Whether public sentiment amplifies the supply shock

Next question

What changes if the rule lands one quarter later?

How it works

Source material in, inspectable forecast out.

The product flow mirrors the real workflow: gather context, run the question through debate, then review the forecast, graph, and reasoning trail in one room.

Step 1

Drop in the context that matters.

Upload reports, notes, decks, transcripts, or long-form text. Prophet turns the material into an evidence map instead of a flat summary.

PDF, markdown, docs, research notes
Claims stay anchored to source material

Step 2

Run the scenario through structured debate.

The room simulates multiple viewpoints, tracks pressure points, and keeps disagreement visible instead of averaging it away too early.

Competing narratives stay separate
Signal shifts become inspectable

Step 3

Read the forecast and interrogate it.

Get a forecast brief, a live graph, and a chat interface for refining the question without losing the reasoning trail.

Forecast, graph, personas, and report in one workspace
Ask follow-up questions without restarting the run

Why Prophet

A summary tells you what was said. A forecast room shows what could happen next.

Use Prophet when the outcome depends on competing narratives, delayed feedback loops, and people reacting to the same facts in different ways.

Best for

Market researchStrategy planningPolicy analysisInvestment theses

Traceability

Every important claim stays inspectable.

You can move from the brief back to the source packet and the graph that produced it, which makes the output usable in real decision reviews.

Disagreement

The room exposes conflict before consensus.

Instead of one polished answer, you see where narratives split, which drivers matter most, and what evidence would change the call.

Iteration

You can keep stress-testing the same question.

Refine the prompt, challenge the assumptions in chat, and preserve the room as the question evolves over time.

Use cases

Bring the question your team keeps arguing about.

Prophet works best when the answer depends on interpretation, incentives, and timing, not just retrieval.

Market and pricing moves

Pressure-test how supply shifts, policy, or sentiment could change the market before you lock the plan.

Narrative and policy risk

Model second-order effects when public interpretation matters as much as the underlying facts.

Research and synthesis

Turn dense source material into a room your team can challenge instead of a summary they have to trust blindly.

Start a room

Stress-test the decision before it becomes expensive.

Create an account, upload the source packet, and turn the question into a room your team can actually work through.

Example questions

What breaks first if the regulation lands two quarters early?

Which narrative wins if public sentiment swings before supply normalizes?

How does the forecast change if the loudest consensus is wrong?