Question 1

What is PRECOGfuture?

Accepted Answer

PRECOGfuture is a strategic forecasting simulation engine. You feed it real evidence — reports, links, documents, notes — and it builds a knowledge graph of the situation, generates realistic stakeholder agents, and simulates how the scenario evolves over time. The output is an auditable scenario report you can interrogate.

Question 2

How is this different from asking a chatbot to predict something?

Accepted Answer

A chatbot gives you one model's single-pass opinion. PRECOGfuture grounds forecasts in uploaded evidence, builds a knowledge graph, simulates named stakeholders reacting across multiple rounds, and exposes sources, agents, events, and verified operations for inspection.

Question 3

What kinds of questions can it forecast?

Accepted Answer

Market entry and product-launch outcomes, policy and regulatory shifts, geopolitical scenarios, crisis response planning, brand and public-affairs reactions, and quantitative time-series baselines where numeric history exists.

Question 4

How much does PRECOGfuture cost?

Accepted Answer

Proposed early-access pricing is Explorer at no charge, Analyst at $149/month, and Institution at $499/month. Plans may change before public release.

Question 5

Can I choose which AI models power my simulations?

Accepted Answer

Yes. PRECOGfuture supports runtime model selection across providers — Gemini, NVIDIA-hosted open models, OpenAI-compatible endpoints, and fully local Ollama models — separately for simulation, graph extraction, and embeddings.

Question 6

Is my data used to train models?

Accepted Answer

PRECOGfuture does not use workspace evidence to train its own models. Hosted model modes send relevant content to the configured provider. A fully local Ollama deployment can keep model processing inside the operator's infrastructure.

Question 7

Does PRECOGfuture produce calibrated probabilities?

Accepted Answer

Not yet. Early outputs are auditable scenario signals for decision rehearsal, not calibrated real-world probabilities. Historical backtesting and calibration are part of the validation roadmap.