Backlog

Next up

In Progress

Done

Main Roadmap

Incrementally evolve Actora’s memory system to improve long-term context retention and retrieval quality. Introduce staged memory strategies, such as selective retention, summarisation, and decay, to ensure the AI remains accurate, scalable, and context-aware as knowledge grows.

In Review

Staged Memory System Improvements

Alberto Zuin

Import projects, tasks, and metadata from <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="chatgpt://generic-entity?number=1">Asana</a> to connect execution work with architecture, governance, and knowledge artefacts, reducing delivery-to-documentation drift.

Asana Project and Task Integration

Automatically ingest boards, cards, and lists from <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="chatgpt://generic-entity?number=0">Trello</a>. Captures lightweight planning artefacts and decisions often missing from formal documentation, linking them to related assets and frameworks in Actora.

Trello Workspace Import

Allow users to rate answers and artefacts for usefulness, creating a feedback loop that helps teams continuously improve knowledge quality.

Knowledge Feedback and Ratings

Provide APIs and webhooks to allow external systems to push updates into Actora automatically when systems, services, or configurations change.

Knowledge Update APIs and Webhooks

Highlight areas of strong or weak documentation and control coverage using visual indicators, helping teams prioritise improvements.

Heat Maps and Coverage Views

Periodically analyse stored artefacts to detect potential duplicates, outdated content, or conflicting documentation. Flags issues for review to keep the knowledge base reliable and current.

Automated Knowledge Hygiene

A browser extension that detects when users are viewing tools such as ticketing systems or cloud consoles and surfaces relevant Actora knowledge automatically, keeping guidance close to where work happens.

Browser Extension for Contextual Knowledge

Enhance the RAG pipeline with agentic retrieval that can inspect full artefact content when needed, not just vector embeddings. The AI can selectively scan complete documents to validate, refine, or correct answers, improving accuracy and reducing false positives caused by purely similarity-based retrieval.

Agentic Retrieval Beyond Vectors

Assess the downstream impact of a proposed change across systems, documents, controls, and teams before implementation.

Change Impact Explorer

Identify recurring questions or topics with insufficient coverage and prompt users to create missing documentation. Optionally generate a first draft based on existing artefacts and conversations.

Planned

AI-Suggested Knowledge Creation

Identify missing documentation, orphaned systems, and unowned decisions based on the current workspace framework.

Knowledge and Risk Gap Analysis

Description:Actora should keep an immutable version history for every artefact (documents, policies, architecture diagrams, manuals, reference links). Each update creates a new version rather than overwriting the previous one. Users can view version metadata (who/when/why), compare versions, and roll back if needed. Versioning becomes the foundation for trusted summaries, review workflows, and conflict detection.Value:Prevents silent knowledge drift, supports auditability (who changed what and when), and enables reliable downstream AI (summaries tied to a specific version, not a moving target).Acceptance criteria (short):<ul><li>Every artefact update creates a new version (previous versions remain readable).</li><li>Each version stores metadata: created_by, created_at, source, checksum, storage pointer, size, MIME.</li><li>API supports: list versions, fetch a version, set current version, rollback.</li><li>Permissions enforced: only authorised users can create versions or rollback.</li></ul>

Completed

Artefact Versioning

Support ingestion of image-based artefacts (JPG, PNG, SVG). Automatically extract and interpret visual content such as diagrams, screenshots, and architectural drawings, linking them to related assets and frameworks to make visual knowledge searchable and usable.

Image Ingestion and Visual Understanding

Provide visibility into how knowledge is used, including top searches, frequently accessed artefacts, and queries that return poor or no results.

Knowledge Usage Insights

Automatically generate concise summaries of long or complex artefacts such as architecture documents, policies, or compliance manuals, improving accessibility for executives and non-technical stakeholders.

AI Summarisation for Long Documents

Assign owners to artefacts and trigger periodic review reminders. Supports simple verification workflows to confirm whether information is still accurate or requires updates.

Knowledge Review and Verification

Query Actora directly from Slack or Microsoft Teams. The assistant can answer questions in context and, when appropriate, proactively suggest relevant knowledge during conversations, reducing the need for manual searches.

Chat and Collaboration Bots

Add optional two-factor authentication to the login flow and enforce email verification for new accounts. Reduces account takeover risk and aligns Actora with baseline corporate security and compliance expectations.

Two-Factor Authentication and Email Verification

Improve the linker by evaluating the semantic and functional context of two artefacts before creating a relationship. Assets are linked only when the relationship makes sense within the chosen framework, preventing incorrect associations such as linking unrelated roles, capabilities, or controls.

Contextual Asset Linking

Enable single sign-on via Microsoft Entra ID and automatically ingest documents from <a target="_blank" rel="noopener noreferrer nofollow" class="link" href="chatgpt://generic-entity?number=0">Microsoft OneDrive</a>. Keeps Actora aligned with enterprise identity and document workflows commonly used in regulated and Microsoft-centric organisations.

Microsoft Entra SSO and OneDrive Auto-Import

Allow users to review, confirm, or correct AI-generated classifications and links, improving trust and long-term accuracy.

actora

Main Roadmap