Question 1

What is PRECOGfuture?

Accepted Answer

A strategic forecasting simulation engine. Give it reports, links, documents and notes. It maps the situation, creates stakeholder agents and simulates how events unfold. The result is a forecast report you can inspect.

Question 2

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

Accepted Answer

A chatbot gives one model's single-pass opinion. PRECOGfuture grounds the forecast in your evidence, then simulates how stakeholders respond as time moves forward. Sources, agents, events and reasoning remain inspectable.

Question 3

What kinds of questions can it forecast?

Accepted Answer

Market entry and product launches, 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

Does it use quantitative forecasting?

Accepted Answer

Not fully. Prophet baselines are available for suitable time-series questions. Broader quantitative tools, including GLM and XGBoost, are coming to the app.

Question 5

How will pricing work?

Accepted Answer

PRECOGfuture uses credits because different runs require different amounts of compute. The app estimates the credit cost before a run starts.

Question 6

Can I choose which AI models power my simulations?

Accepted Answer

Yes. You can choose models from OpenAI, Anthropic, Google, DeepSeek, Moonshot AI, Qwen and Mistral, including GPT-5, Claude Opus and Sonnet, Gemini 3, DeepSeek, Kimi, Qwen and Mistral models.

Question 7

Is my data used to train models?

Accepted Answer

PRECOGfuture does not use workspace evidence to train its own models. A fully local Ollama deployment is possible and can keep model processing inside the operator's infrastructure.

Question 8

Does PRECOGfuture produce calibrated probabilities?

Accepted Answer

Not yet. The results are structured scenarios and evidence-backed signals, not guaranteed outcomes. Historical backtesting and calibration are part of the validation roadmap.