The Jube roadmap is shaped by production experience across a significant number of production deployments. What follows reflects real operational feedback from compliance teams, implementation partners, and co-delivery engagements — not theoretical product planning. Items are sequenced by impact and dependency, not by ambition.
The roadmap is published in plain terms. No commitments are made to specific release dates beyond the releases described below. The platform evolves continuously and this page is updated to reflect current direction. Between formal releases, the platform continues to evolve in response to the requirements of the current corpus of production clients; changes of general application are merged continuously and will be present in 0.1.0 onwards, as shipped.
Version 0.1.0 — Consolidation Release — July 2026
Version 0.1.0 is the first formal tag of the Jube platform. It represents a deliberate consolidation of production-proven capability across a significant number of production deployments — tightening existing functionality, closing identified gaps in operational completeness, and establishing a stable, versioned baseline for all subsequent development.
This is not a feature release in the conventional sense. The feature additions included are directly informed by production client requirements. The majority of the work is hardening, integrity, and operational completeness rather than new capability.
Security and Authentication Overhaul A comprehensive platform hardening including hybrid JWT and API key authentication across all endpoints, parameterised SQL throughout, AES encryption hardening, password generation improvements, and HTTP security header injection. Breaking changes to Invoke endpoint authentication are included; all endpoints now require authentication.
Performance and Resilience Under Load Significant work on connection pool management, Redis Polly retry policies, Sentinel failover support, cache warm jitter and batching, and resolution of a long-standing connection exhaustion root cause in the PostgreSQL connection wrapper. Benchmarks on development hardware show substantial headroom against production throughput requirements.
UTC Standardisation Full stack standardisation to UTC timestamps.
Preservation Corrections to import and preservation logic covering edge cases identified in production and role migration support.
Cache Architecture Optional offload of HSET operations from Redis to PostgreSQL, a new CacheSetHash table with covering index for index-only scan performance, and an abstraction interface over the Redis cache layer to support future backend substitution.
Tagging Infrastructure Tags promoted to a data structure with their own table, archive integration ensuring tagged records are reflected consistently across case records, and tag management within the Case Key Journal and a dedicated Transaction Report page.
Idempotency Idempotency for Time to Live Counter incrementing at the transaction level, and Case Creation and Notifications at the transaction and activation rule level.
Manual Case Creation via Transaction Report Page The ability to retroactively create cases on the basis of results in the aforementioned Transaction Report page. Structured case creation directly from transaction and alert context, with full audit trail and journal from point of origin. Closes the workflow loop between detection and investigation.
Sampling and Analytics Archive sampling page enabling date-range and percentage-based sample extraction from the transaction archive, with JSON flattening to CSV for downstream analytics use.
Date Expiry on Suppressions and List Entries Time-bounded suppression and list management, ensuring that entries age out automatically rather than accumulating indefinitely. Addresses a common operational hygiene problem in production deployments and supports defensible compliance posture.
Add to List from Case Journal The ability to add entities, values, or identifiers directly to a List from within the Case Journal, without leaving the investigation workflow. Reduces analyst friction and accelerates response.
IP Intelligence Dataset Integration Integration of a proprietary IP intelligence dataset built from multiple corroborating public sources, cross-validated to produce reliable attribution at the country, ASN, and network type level. The dataset is maintained with a specific compliance focus — prioritising attributes that can be verified with confidence over those that carry an impression of precision without the substance to support it. City-level geolocation is deliberately out of scope; the dataset is designed around what remains accurate and defensible under compliance scrutiny, not what looks compelling in a demo.
Cluster and High Availability Support of fully active-active clustered deployments with no single point of failure, suitable for primary and Disaster Recovery configurations. The reference cluster topology comprises multiple Jube application nodes, a Patroni-managed PostgreSQL cluster with automatic failover coordinated via etcd, and a Redis master-replica setup with Sentinel for near-real-time failover detection. The platform is designed to run in this configuration, typically in Docker Swarm, though Kubernetes is equally supported, in production; horizontal scaling of application nodes requires no architectural changes.
Version 1.0.0 — Tests and User Experience — March 2027
Version 1.0.0 is the full platform release. The backend capability established in 0.1.0 is complete; this release delivers the user experience to match. It is a comprehensive UX investment.
Integration Test Coverage A structured integration test suite focused on the invocation pipeline and model construction via the API. Test cases will be built around real integration scenarios — exercising the full path from invocation through model execution — rather than controller-level unit coverage. Controllers are covered only to the extent that they participate in meaningful end-to-end scenarios. The test suite will be designed to support continuous integration and protect feature velocity as the platform matures: changes to core pipeline behaviour are caught early, and new capability can be delivered with confidence against a stable, verified baseline.
Platform UI Migration — Daisy and Vue The full platform UI will be migrated to Daisy UI and Vue, replacing the current jQuery-based implementation. The migration will deliver a modern, component-driven interface with consistent design language across all platform areas. The configuration interface — covering rule management, model configuration, entity setup, and platform administration — will be migrated as a direct translation of existing functionality.
Visualisation The visualisation layer will remain SQL-led and largely unchanged in its underlying approach. The port to Chart.js will replace the current rendering implementation while preserving the query-driven model. The primary UX improvement will be a move away from requiring analysts to author raw JSON initialisation blocks: charts will instead be defined through templated configurations with user-defined parameters and series definitions, keeping the flexibility of the existing approach while substantially reducing the technical burden of chart authorship. More broadly, visualisation will no longer be confined to the Case Management and Visualisation Directory contexts — chart and query output will be available across a wider range of platform areas, bringing data-proximate rendering to the workflows where it is most useful.
Case Management Redesign The case management interface will be rebuilt from the ground up with serious usability at its core. The existing interface was designed for functional completeness; the redesigned interface is being designed for the compliance analyst who lives in it all day. Workflow, information architecture, and interaction design are all in scope.
AI-Assisted Case Automation LLM-driven automation within the case management workflow via Semantic Kernel, supporting narrative generation, case summarisation, next-action recommendation, and structured reporting. Designed to reduce analyst burden on routine documentation tasks and accelerate case throughput without reducing quality or auditability. Case management functions otherwise available via API are exposed through the same integration layer and invocable given prompts via Semantic Kernel.
Vector Similarity Analysis Embedding-based similarity analysis across case history, enabling the identification of structurally similar cases across time, entity, and typology dimensions. Intended to support pattern recognition at scale, typology development, and the surfacing of related activity that rule-based approaches may not connect. Particularly relevant to complex layering and integration-stage AML typologies.
Principles
Open source first. Jube is and will remain open source under AGPLv3. The roadmap reflects the platform’s mission to make serious AML and fraud detection capability accessible. Commercial arrangements exist to support sustainable development, not to gate functionality.
Production informed. Every item on this roadmap has been validated against real deployment experience. Nothing here is speculative.
No roadmap theatre. This page will be updated when direction changes. Items will be removed if they are deprioritised. The roadmap exists to communicate genuine intent, not to market a vision.
Last updated: May 2026