Intelligence product case study

Trial Radar demonstrates an end-to-end intelligence product: ingest public registry signals, structure them with taxonomies and schemas, enrich entity profiles, and distribute recurring briefs — the same playbook premium research publishers use for enterprise data products.

Architecture

ClinicalTrials.gov API
        ↓
  Normalize + hash diff
        ↓
  Event taxonomy + severity
        ↓
  Entity schema (trial, trial_change, snapshot)
        ↓
  ┌─────────────┬──────────────┬─────────────┐
  │  Landscape  │ Entity       │  Recurring  │
  │  dashboard  │ dossiers     │  briefs     │
  └─────────────┴──────────────┴─────────────┘
        ↓
  Reports · Digests · Email alerts

Job requirement mapping

Extract & structure intelligence

Change detection pipeline classifies registry diffs into typed events (status, enrollment, geography, phase, protocol). Entity dossiers and activity threads package signals into analyst-ready formats.

Taxonomies, tagging & schemas

Event taxonomy with severity scoring (critical / notable / minor), normalized entity schema, and point-in-time snapshots for version history.

Enrich & validate against public sources

All entity records cross-referenced with ClinicalTrials.gov API v2. Content-hash deduplication prevents false positives; snapshots validate state after each sync.

Recurring data feeds

Daily Intelligence Brief (editorial narrative from structured signals), scheduled email digests, and weekly competitor briefings via cron jobs.

Commercial product thinking

Products tiered for enterprise buyers: sector briefs, entity dossiers, change feeds, and custom watchlist monitoring. Built for corp dev, investors, and advisory firms.

Event taxonomy

Event typeSeverityExample signal
Status changeCriticalRecruiting → Completed
Phase advanceCriticalPhase II → Phase III
Enrollment updateNotableTarget enrollment revised
New registrationNotableNew entity registered
Geography expansionMinorSites added in new country
Protocol editMinorIndication or summary updated

Validation & enrichment

  • SHA-256 content hashing skips unchanged records — reduces noise in change feeds
  • Point-in-time snapshots captured after each detected change for audit trail
  • Every entity record links to source NCT ID on ClinicalTrials.gov for verification
  • Parallel count queries validate aggregate landscape metrics against live API

5-minute demo script

  1. Landscape — Daily Intelligence Brief + sector KPIs
  2. Entity dossier — sponsor profile + activity thread
  3. Open any NCT ID — entity activity thread with snapshots
  4. Activity threads — per-study change history on the landscape
  5. Settings — watchlists and alert preferences

Portfolio demonstration of intelligence product development · View product catalog