untouched
Raw source
Every file lands in an insert-only registry, hashed with sha256. Originals are never modified — the record is append-only by construction.
Provenance-first intelligence
Your archive knows
the story.
Ask the body of work you already built. StoryIQ connects your sources, shows the threads between them, and keeps every answer attached to the passage that earned it.
app.storyiq.net / property-play / ask
ABWhich ideas have shown up in my reporting—but never became a story?
StoryIQSOURCED FROM YOUR ARCHIVE
Three threads recur without a dedicated piece. The strongest is the widening gap between conversion announcements and permits —followed by insurance costs quietly replacing rates as the migration constraint .
You already have five source relationships and two unpublished transcripts that can move both ideas forward .
1 · article2 · transcript3 · notes
Ask a follow-up about your archive…↑
Benefits
Every claim, accounted for.
StoryIQ gives you real intelligence without the data overload — and never asks you to take its word for anything.
Grounded extraction
Every field is lifted straight from the document, pinned to the exact verbatim passage it came from.
Resolved entities
Mentions collapse into stable entities by ticker, alias, and fuzzy match — a model judges only on real ambiguity.
Human in the loop
Nothing reaches the graph until a person reviews, corrects, and signs off. Your edits become the canon.
Search by meaning
Completed documents are embedded, so the graph answers by meaning — grounded in the version you verified.
How it works
Five layers, each citing the one below.
machine-true
Deterministic facts
Send dates, message-ids, dedup hashes, clean reading paragraphs — parsed, not guessed. The same input always yields the same facts.
evidence-pinned
Grounded extraction
An agentic parse reads each document against your schema and returns structured fields — each one carrying the exact passage it came from.
resolved
Canonical graph
Mentions resolve to stable entity ids by ticker, alias, and fuzzy match. Relationships become temporal, cited edges in one graph.
reviewable
AI enrichment
Cross-document edges, merge suggestions, and cited research — emitted as proposals a human accepts or rejects. Never written silently.
And nothing is ever overwritten.
Each layer references the one beneath it by id — so you can always walk a claim back down to the byte it came from.
See the graphThe graph
Click a node, read the passages.
People, projects, topics, articles, and episodes resolve into one graph. Every node carries the exact sources it was drawn from — nothing floats free of its evidence.
grounded view · 10 entities · 13 edges
- Person
- Topic
- Project
- Episode
Marcus Liang
person · 14 appearances
aliases · Marc Liang · M. Liang
SOURCES
- Ep. 39transcript · 14:22
- Apr 8Vacancy Watch · ¶3
Idealized view of the full-page graph. In the app, selecting an entity opens its source list and first-seen / last-seen timeline.
Specifications
The difference is the receipts.
CapabilityStoryIQGeneric RAGManual review
Evidence span for every claim yes no~ partial
Entity resolution to stable ids yes~ partial~ partial
Human review gate before canon yes no yes
Drift auditing after schema edits yes no no
Grounded-only by default yes no yes
Idempotent re-ingestion yes no no
Search by meaning (embeddings) yes yes no
Pricing
A research desk that grows with the story.
Start with one archive. Add the room, controls, and throughput your team needs. Prices are managed in Stripe and can change without a product rebuild.
Independent creators
Solo
$24per month
A focused intelligence desk for one body of work.
- ✓1 active workspace
- ✓2 connected sources
- ✓Sourced AI answers
- ✓Knowledge graph
Editorial teamsMost popular
Desk
$79per month
Shared context and faster reporting for the whole desk.
- ✓5 active workspaces
- ✓Unlimited connected sources
- ✓Team invitations
- ✓Priority ingestion
Growing organizations
Newsroom
$199per month
Governance and scale for a newsroom-wide intelligence layer.
- ✓Unlimited workspaces
- ✓Advanced permissions
- ✓Audit history
- ✓Priority support
Placeholder monthly prices shown. Your live Stripe Price objects are the source of truth at checkout.
Get started
Put your archive to work.
Open the Reading Room and start asking your own archive — every answer cited to the passage it came from. Or request access and we’ll set up your workspace.
Open StoryIQ Continue at app.storyiq.net
Or request access