ARCHITECTURE

Galaxy is a single-tenant solution (each platform dedicated to one client) that can be deployed at the client's premise, on a client's cloud environment or operated as a service on Dalet's cloud account. Multi-site and hybrid on-premise/cloud operation is supported by machine resource and service location awareness. All this adds up to an intelligent system suitable for operation over multiple geographic locations. There is no human resource location awareness support though. While Dalet Galaxy is built on open standards with SOA-compliant Web Services APIs and has a unique distributed architecture with multiple layers, each dedicated to a level of functionality and service, it is not fully micro-services based (although this is currently being addressed through Dalet Pyramid, Dalet’s next-generation news production solution, which is fully microservices-based) or containerised/dockerized, which may make it more time consuming to deploy and support than other platforms that are. In terms of metadata store, Galaxy relies on a relational database and on a solr Search index. The search index allows delivering responsive answers to queries while the relational database allows building advanced data models.

​

INTEGRATION

Dalet Galaxy has extensive integration capabilities built on a framework of user-configurable plugins, scriptable nodes in their workflow design tool, and a web services RESTFul API (although no SDK for self-build plugins). It supports all standard AV containers and formats (including MXF and IMF) and the main exchange formats (BXF and MOS). Galaxy has proven integration with most on-premise machine resource types (on-line/near-line/deep storage, accelerated file transfer, AQC, transcoders, audio processing and video processing), cloud machine resource types (online/archive storage, AQC, AI/ML, web CMS and social media), and communications systems (MS Teams). Cloud transcoders are supported natively with Dalet AmberFin Cloud Transcoder and via the workflow engine scripted stencils for other vendors.[MF1]   Missing from their current integrations is on-prem AI/ML, and cloud linear playout systems - all of which could possibly be implemented on request. Integration with other intelligent systems is also strong, including rights management systems, scheduling & traffic systems, hierarchical storage management systems (HSM), media asset management systems (MAM), video on demand systems (VOD) and newsroom computer systems (NRCS). Missing from their list is integration with customer relationship management systems (CRM), eCommerce systems and facility management systems. Of note, if you have legacy infrastructure that you want to integrate with, Galaxy supports SDI to file ingest, file to SDI visual QC, and file to SDI playout. It also is able to integrate with SDI glue machine resources (routers, switchers...), has VTR and Flexicart control, and has proven integration with linear playout systems and vault library systems.

​

AUTOMATION

Dalet Galaxy includes a very aesthetically pleasing graphical workflow design tool that claims to be BPMN 2.0 compliant. With this, you can design from simple single linear path workflows to complex multi-path workflows, including launching workflows from within workflows and writing your own scripted nodes. User decision points can be added, and workflows are real-time metadata responsive. Galaxy supports the automated allocation, prioritisation and real-time modification of machine actions, including load balancing. Automated human resource task setting is supported by group and user pools, and timed tasks can be set. Galaxy does not support automated machine resource scheduling, nor the spinning-up of additional cloud resources on demand. It is not location-aware, cost-aware or duration aware, which could arguably limit the efficiency of some automated workflow applications. Monitoring and reporting of automated activities appear to be a strength in Galaxy, including the allocation of tasks, queue monitoring, time taken to complete tasks, and automation progress, presentable as either text, UI components and/or workflow graphics. Configurable exception handling capabilities can automate the notification of appropriate users and user groups or third-party systems of specified events.

​

METADATA

Dalet Galaxy enables custom metadata design and management to be performed through technical configuration and configuration tables but does not have an easy-to-use drag and drop UI that some other solutions offer. It supports media asset, business object and other file type metadata definitions and maintains asset, object and file metadata history. It does not support resource usage or rights metadata. For media assets, Galaxy supports temporal metadata, including single timecode point markers/tags and timecode in and out durations, each with its own metadata definitions. Object hierarchies, asset genealogy and asset collection are supported. Spatial metadata is not supported. Asset download control, asset expiration dates, limited use restrictions and archive restrictions are supported. An asset to contract linking, version control and notifications, edit usage rights, access revocation and sensitive material lockdown is not supported. Metadata on the users, user groups and user task history are supported, as well as machine action history (events). Galaxy includes a chat capability with a retained history, and emails sent via the BPM can be customized. Galaxy supports a host of metadata access and entry control functionality, including customisable taxonomies and thesaurus, field-level search/index option, field-level linking to backing store, customisable keywords and tagging and customisable entry validation. It does not support the limitation of metadata entry by role or user group.

​

USER EXPERIENCE

The Dalet Galaxy user interface reflects a modern look and style and its media operation feature-rich. It has a consistency of theme and layout, components, and taxonomy, with the ability to remove features by role. The UI has some limitations, including the inability to customise the table headers or set customisable filters. Content search capability within Galaxy seems to be one of the best on the market, from basic search with auto-complete to advanced search with Boolean layers. Searches can be filtered by asset types and attributes, business object attributes, metadata fields and tags. A built-in configurable thesaurus, configurable taxonomies and tags and AI-driven recommendation all assist the operator to efficiently locate the content they need. For collaborative working, the Galaxy UI provides contextual To-Do lists (automatically set, for individual users or user pools), an in-app chat, and configurable timelines with comment entry. Dalet also offers Android and IOS apps, further strengthening the ease of collaboration. From a creative perspective, Galaxy offers more tools than many of its competitors. Galaxy includes a non-linear editor (NLE) that is more capable than the cut-cut capabilities in most platforms. It also supports voice-over recording and editing, audio track selection and swapping, and caption generation, entry and viewing.

​

ANALYTICS

Dalet Galaxy has both fixed charts built into its UIs and support for custom charts through integration with third-party apps (like Kibana). There appears to be little in the way of automated reporting capability, UI component representation or other graphical visualisations or representation. The presentation of performance metrics appears strong in Galaxy, including the ability to present task durations and throughput by user and group, and filter automated jobs that have failed, completed or in-progress. Missing here though is the ability to analyse bottlenecks - queue times for machine actions or human tasks, or the time to progress a program to air. Galaxy provides some quality metrics analysis capability, including integration with leading third-party AQC products, QC failure analysis by the content supplier and issue type, metadata entry checking and analysis and platform issue tracking. Usage metrics analysis is limited to machine resource usage by time, with no analytics of machine usage by location or capacity, nor human resource usage (by time, capacity or location), nor asset usage (by supply chain stage, distribution or asset usage versus rights). Galaxy supports no financial analytics, so no cost tracking, cost dashboards, billing tracking, analysis of cost versus budget etc.

​

SECURITY

A strong commitment to security is shown by Dalet with Galaxy scoring well in nearly all areas that we considered. With Documentation and Testing, Dalet confirmed that they had met all the requirements for policy, audit and recovery, and their secure coding practices seem to follow best practice (although it should be noted that we do not actively validate either of these areas). Regular penetration tests are run by a 3rd party specialized company on each Dalet Galaxy five new version to identify a potential security breach. Secure centralised authentication is provided with the principle of least privilege and authenticates first, authorise next policies adopted. Secure password with a recovery mechanism is supported, but not multi-factor authentication or biometric authentication. Logging and forensics capabilities include logging timestamps with source IP, an extensible logging framework, and exception logging, but no automated security breach alerts. Within Galaxy, security breach prevention includes SQL intrusion prevention (query parametrisation, object query limitation, but not database management system configuration), input validation (server-side) and API security measures. 

​

ARTIFICIAL INTELLIGENCE

Dalet Galaxy has proven integration with Microsoft Video Indexer, Speechmatics and Google AI and Dalet have its own in-house AI capability, Media Cortex, which is an AI service platform designed to leverage cognitive services and machine learning across content operations and business. Galaxy has not yet been productized integration with Amazon Rekognition, IBM Watson or any of the AI aggregators (Valossa, Veritone, GreyMeta), although such integration could potentially be implemented if required. Combining this AI framework with 'smart workflows', Galaxy can automate facial recognition tagging, visual text extraction (OCR), language detection, language conformance detection, voice-to-text transcription, captioning, and multi-language translation. It can also automatically identify logical scene boundaries and perform legal compliance and censorship checking. AI is also applied in Galaxy for content analysis and enhanced search, including context detection, intelligent discovery and contextual recommendations. AI functionality has yet to be extended with Galaxy to emotion, age, logo, object and voice detection than some other platforms reviewed offer (many of these could presumably be added by Dalet though if they are already supported by the Microsoft AI or Google AI).  AI has not been applied in Dalet Galaxy for any form of technical processing (colour detection, brand accuracy detection, colour correction),  operations management (machine or human resource optimisation, rights provisioning, on-demand application or service sourcing etc), advanced content creation (automated storytelling compilations or highlight packages) or intelligent placement (linear schedule creation, VOD publishing and grouping or ad placement).