From market-moving news to explainable stock signals.
Hurd Stocks is being built to collect relevant market information, connect it to affected assets, score likely significance, and show users the evidence behind each output.
Discover
Pull relevant market information from multiple public sources, not just a single feed.
Normalize
Persist source material in a durable form so analysis and reprocessing do not depend on refetching the original page.
Link
Connect stories to companies, sectors, and related assets with explicit evidence and confidence routing.
Deliver
Return ranked views, alerts, and explanation layers that help users decide what deserves attention first.
Asset-first workflow
The product is designed around the question of what matters to a stock, sector, or watchlist, rather than around publisher-oriented headline browsing.
Search as a retrieval layer
Full-text indexing helps with retrieval and investigation, but the durable source of truth remains the stored article snapshots and versioned outputs.
Confidence-aware outputs
Signals should not feel equally certain. Hurd Stocks is being built to expose confidence and boundaries alongside the interpretation itself.
Replay-friendly design
The pipeline is being designed so linking, scoring, and evaluation can be rerun as the models and dictionaries improve.