Hiring Signals | Intelligence Desk

Company	Signal	Score	State

Select a company

Choose an account to inspect, edit its evidence and record next actions.

Add candidates manually or import a JSON database. No interview transcripts or personal candidate information are included in this starter app.

Candidate-to-company fit (preliminary)

The matcher uses role and sector keyword overlap plus signal score. It is a triage aid, not evidence of a vacancy or candidate availability. Verify geography, pay, achievements and consent before contact.

Account-led BD queue

Assign a commercial stage, next action and stakeholder for each account. This is an operational queue, not a job board.

Account	Signal	Stage	Decision maker	Next action

Headline discovery inbox

Import discoveries.json generated by the scheduled scanner. These are public search results that need source and org-chart verification. No headline is automatically certified.

Research operating procedure

1. Scan issuer news, filings, authoritative trade outlets, accessible public posts, leadership pages and live vacancies.

2. Record a dated primary source. Classify completed, announced, pending or rumoured events.

3. Stack at least two independent facts. Score structure (0–2), complexity (0–2), gap (0–2), economics (0–2), budget (0–1), and mandate (0–1).

4. Investigate the named role owner, reporting lines and an actual unfilled capability gap. Distinguish verified event from confirmed funded hiring.

5. Match credible candidates and create decision-maker outreach. Log response, objection, next action and outcomes.

6. Recheck changes and deduplicate by company + event + role. Never re-alert unchanged stories.

What this standalone app can and cannot do

Works now: seeded foodservice accounts, editable scoring, saved notes, candidate intake, basic fit ranking, JSON backup/restore. Data is stored in this browser only.

Needs integration: live company news, RSS/Google Alerts, Metaview transcripts, CRM accounts, automated source verification, multi-device sync, scheduled database updates and email alerts.

Production architecture: Postgres/Supabase database + scheduled ingestion workflow (n8n/Make or hosted worker) + LLM classification + source audit trail + app frontend. LinkedIn content must be gathered using authorised methods only.