Timeline Domain

Overview#

An investigator is reviewing a complex financial fraud case spanning eighteen months and four jurisdictions. Rather than scrolling through thousands of raw records, she opens the investigation timeline, where AI has already extracted and ordered the key events: account openings, wire transfers, company registrations, and phone contacts, each on its own track, each linked to the underlying evidence. She spots an overlooked three-day gap that turns out to be a money mule drop. That kind of AI-assisted chronological clarity is what the Timeline domain delivers.

The domain provides multi-track investigative timeline management with AI-powered event extraction, temporal relationship inference, and automatic importance scoring. Investigators can visualise case chronology, link events across multiple evidence sources, and generate timelines automatically from investigation data stored in PostgreSQL.

Key Features#

Multi-Track Timelines: Organise events across multiple parallel tracks (such as communications, financial, physical movement, and digital activity) for comprehensive chronological analysis of complex investigations.
AI Event Extraction: Automatically extract events from evidence documents, communications, and other data sources using AI analysis, reducing manual timeline construction effort.
Temporal Relationship Inference: Identify causal, sequential, and concurrent relationships between events using AI analysis to surface connections that may not be immediately apparent.
Importance Scoring: Automatically score the significance of timeline events to help investigators focus on the most important moments and avoid information overload.
Event Management: Create, edit, and organise timeline events with rich metadata including dates, descriptions, categories, evidence links, and entity associations.
Automatic Generation: Generate investigation timelines automatically from available case data including evidence timestamps, communication records, and financial transactions.
Timeline Export: Export completed timelines for inclusion in reports, court presentations, and briefing documents.

Mermaid Diagram#

Use Cases#

Law Enforcement: Build a comprehensive chronological view of a criminal investigation spanning multiple evidence types and data sources, ready for court presentation with evidence links supporting each entry.
Financial Crime: Identify temporal patterns and correlations between financial events across different tracks that may reveal coordinated money laundering or fraud activity.
Defence Intelligence: Create clear, well-organised timelines of adversary activity for briefing command staff, with classification-appropriate access controls on each track.
Corporate Investigations: Share multi-track timelines with legal and compliance teams to enable collaborative analysis of complex cases involving many concurrent business activities.

Integration#

The Timeline domain supports chronological analysis across the platform:

Investigation Management: Timelines are associated with their parent investigations.
Evidence Management: Timeline events link to supporting evidence items.
AI Services: Event extraction and relationship inference use AI analysis.
Reporting: Timeline exports feed into investigation reports.
Storyboard: Timeline events can be incorporated into narrative storyboards.

Open Standards#

GraphQL (June 2018 specification): the entire Timeline domain API, including event queries, mutations for creation/update/deletion, and AI extraction operations, is exposed as a typed GraphQL schema via Strawberry.
ISO 8601 / RFC 3339: all timeline event timestamps, viewport bounds, and duration values are stored and exchanged as UTC-normalised ISO 8601 date-time strings with explicit timezone offsets.
JSON (RFC 8259): event metadata, extra_data payloads, temporal relationship records, and vector embeddings are serialised as JSON for storage in PostgreSQL and transport over the API.
OAuth 2.0 / JWT (RFC 6749, RFC 7519): timeline endpoints are protected by the platform's JWT-based authentication layer; the IsAuthenticated permission class validates bearer tokens on every query and mutation.
pgvector (IEEE floating-point vector extension for PostgreSQL): event similarity search uses cosine distance over 768-dimensional embeddings stored in a pgvector column, enabling nearest-neighbour retrieval directly in the database.
W3C WCAG 2.2: the multi-track timeline visualisation and interactive controls are expected to meet WCAG 2.2 accessibility conformance for court-presentable and public-sector use cases.
W3C ARIA (Accessible Rich Internet Applications): dynamic timeline tracks, playback controls, and filter panels use ARIA roles and live-region attributes to support assistive technologies.

Last Reviewed: 2026-02-05 Last Updated: 2026-04-14