Datahub is the record your teams work from.

One sourced profile per person and organisation: roles current and previous, positions by issue, relationships, and the date it last changed. New stakeholders are drafted by the models from the record and reviewed before they publish. Duplicates from Outlook and spreadsheets resolve into one.

Roles, current and former

Public role changes are detected and applied, with the history kept. A regulator turned adviser is read as both.

Positions by issue

Where they stand on each issue in scope, with the statement it came from and its date.

Lists and watches

Priority lists, programme groups and watches, shared by team or kept to one.

Landscape, influence and relationship views drawn from the typed relationships in the record, with your own relationships marked and the people who moved since the last review lifted out.

Typed, sourced relationships

Formerly at, member of, funds, advises, reports to: each tie typed, sourced, and marked verified or inferred.

Layers

Switch a competitor, a programme or a region on and off to see what is shared.

Your relationships marked

The people your teams know are flagged on the map, with the last contact.

More than two million items a week from nine source types are cleaned, de-duplicated and classified by twenty or more models before anyone reads them: tagged to your issues, scored for sentiment towards each stakeholder named, resolved to the actors behind them, and ranked for relevance to your workspace. What reaches the feed is the fraction that bears on your mandate, with its source and time.

Your issues, your frameworks

Classifiers trained on examples our analysts label with you. Items are sorted the way your team would sort them.

Resolved to the actors

Every mention is matched to a profile. A story about the minister lands on the minister, whatever name it used.

Saved filters and alerts

Any view becomes a standing watch, with an alert when it spikes, a new voice appears or a position shifts.

Nine source types

News and wires, social, academic, government and regulatory records, NGO and campaign material, company filings, and your own licensed feeds.

Sentiment and narrative are tracked over time against your framework, by audience and language. Reach a spike and open what drove it that day, down to the items and the stakeholders behind them.

Against your framework

Reputation dimensions, risk categories or material issues, configured as your teams report them.

By audience and language

Read separately by the audiences that matter and the languages they speak, not as one blended score.

Drivers, not just scores

Reach a spike and open the stories, the sources and the stakeholders that drove it that day.

Narratives identified at scale

Context-aware AI classifiers label your entire signal universe against the narratives you watch, your issue framework or any taxonomy you choose. The most granular narrative analysis available.

Meetings and emails with known stakeholders are captured from connected calendars and inboxes. Notes, commitments and attendee lists go against the same profile, visible to the teams allowed to see them.

Captured from calendar and inbox

Meetings and emails with known stakeholders are logged from Outlook and Google without re-keying.

Notes and commitments

What was said and what was promised, against the person, with a date to come back to.

Coverage by team and market

Engagement against priority lists reported by team and market, with the group view above.

A question becomes the filters it implies, shown so you can adjust them, and the answer links to every item it came from. Each user is answered from what that user can see.

Filters you can see

A question becomes the filters it implies, shown so you can adjust them and ask again.

Sources attached

Every answer links to the items it came from. Check it where you asked.

Answered per user

Each person is answered from what they are allowed to see.