Our products and open research
Forecast Library
A permanent, inspectable memory for forecasting science. Built on Open Forecast Receipt for forecasts across domains, methods and publishers.
Forecast Library and iPulse AI are distinct products. The Library's broader direction includes weather, demand, scientific and other measurable forecasts from people, models and agentic systems. iPulse AI focuses on financial markets and is the Library's first live publisher; its historical records connect back to the relevant market research.
Forecast Library · A Future Edge Group project
Current research. A permanent record of what came before.
Forecast Library is a general project for forecasting science: preserving inspectable evidence across subjects, methods and publishers. Its first live publisher is iPulse AI, our financial market research platform. Their connection lets readers inspect original market forecasts and return to current AI Consensus without making the Library dependent on a financial taxonomy.
Inspect the original prediction
Follow an iPulse AI historical forecast to its exact record, with the prediction, forecaster and time boundaries preserved.
Verify the evidence
Check that the sealed receipt has not changed. Integrity verification and optional blockchain timestamps are separate from forecast accuracy.
Return to current research
Move from the historical record to the matching asset's current iPulse AI research and AI Consensus.
From an open-source experiment to public forecast evidence
Open Forecast Receipt started with a practical question: can someone inspect exactly what an AI predicted after the conversation, model or website has changed? The first demonstration sealed an individual forecast and showed how changing one value breaks its integrity check. Forecast Library makes those records browsable and connects them to the original research.
The first complete public collection contains 4,511 individual iPulse AI forecasts across 376 subjects. These are published forecast records, not a claim of measured predictive accuracy or 4,511 independent models.
Watch the original Open Forecast Receipt demo on YouTube ↗What is Forecast Library?
Forecast Library is Future Edge Group's public library of forecast evidence. Each published forecast has a permanent identity and a sealed receipt that can be inspected and verified. iPulse AI is its first publisher and market research is its first live use case.
How is it different from iPulse AI?
iPulse AI is our Open Agentic Investment Research Platform for exploring current AI Consensus, individual forecasts, methodology and market context. Forecast Library provides the preserved record behind historical forecasts. A receipt links back to the corresponding asset research, keeping present research and original evidence connected.
What is Open Forecast Receipt?
Open Forecast Receipt is the open-source schema and verification toolkit underlying the library. It describes what was predicted, who or what produced it, when it was generated, the relevant information boundaries and how the payload is sealed. The standard, the public library and optional blockchain attestations are separate layers.
What does verification prove?
Receipt verification checks whether the payload matches its recorded cryptographic digest. An independently verified blockchain attestation can add evidence that a digest existed by an anchoring time. Neither proves accuracy, sound reasoning or investment suitability. A retrospectively created receipt does not establish that the forecast was sealed at its original generation time.
What happens if a forecast is corrected?
The original evidence remains identified by its own receipt digest. A correction needs a new record and explicit lineage. Existing published record addresses must continue to identify their original evidence, independently of page layout, asset names or model labels.
How does this connect to agentic workflows?
The project began as an open-source initiative to make forecast evidence portable. The ambition is for research agents to exchange receipts, preserve provenance through Agent-to-Agent handoffs and attach evaluations when outcomes become known. The library and verifier are available today; automated A2A workflows and general evaluation services remain future work.
How can I submit a forecast?
Contact support@ipulseai.com for manual review. Include the forecast, subject, measurable target, generation date and information boundaries, methodology or forecaster identity, and confirmation that you may publish the supplied material. Do not send credentials or confidential information. There is no self-service submission or payment requirement.