PAGE CONTENTS
Objectives
GAI aims to streamline and elevate satellite gateway service management by turning disconnected corporate information into coherent, actionable knowledge. The project will:
- Integrate and harmonise data from B/OSS tools, CMDBs, communication systems, gateway and service documentation, and selected OT sources such as weather feeds.
- Construct a unified knowledge map and holistic indicators for deployment, capacity, and performance, grounded in enterprise policies and GAI’s governance model.
- Apply background capabilities (asset consolidation, noise reduction, anomaly detection, event correlation, case intelligence, policy consolidation, and others) to reduce operational friction and improve situational awareness.
- Expose unified, intelligent features such as AI Companion, Dynamic Capability Map, Athena, Predictions & Proactivity, and Smart Service Workflows, to help operators, engineers, and managers plan, execute, and govern services with confidence.
Challenges
Satcom organisations operate across multiple disconnected systems, teams, and lifecycle phases. Their ITSM, O/BSS, communications, documentation, and OT environments frequently lack integration or common policy alignment.
This fragmentation causes:
- Reduced visibility and unclear global status of gateway assets and services.
- High operational noise, duplicated work, and poor knowledge reuse.
- SLA exposure due to slow detection, slow decision-making, and reactive workflows.
- Lack of end-to-end governance from design to operations.
GAI must overcome these challenges by harmonising heterogeneous data, enforcing policy consistency, and delivering trustworthy insight without disrupting existing enterprise tools or processes.
System Architecture
GAI’s architecture follows a modular, data-centric design that integrates Data Ingestion, Transformation, Cloud Storage, and AI-driven Features. It connects to diverse corporate ecosystems – B/OSS, CMDBs, communication platforms (Teams, Outlook, Slack), and OT sources (e.g., weather APIs), through secure APIs and RPA pipelines.
Within the GAI Engine, data undergoes consolidation and transformation via core modules: Asset Consolidation (identification, tagging, asset identity resolution), Noise Reduction (deduplication, filtering, linking), Anomaly Detection and Event Correlation (tailored reporting and pattern recognition), and Holistic Indicators (asset and case KPIs). These feed into Case and Workflow Management, enabling chronological case reconstruction, prioritisation, and service consolidation. Policy Consolidation enforces corporate governance rules, while Quality Assurance modules apply confidence and sentiment scoring for continuous system improvement. All processed data is stored in the Azure Cloud using JSON, SQL, and vector databases, exposing insights via AI APIs.
The Feature Layer, comprising Athena, Dynamic Capability Map, Smart Workflows, Predictive Analytics, and AI Companion translates technical intelligence into actionable user outcomes.
This scalable architecture ensures data integrity, modular extensibility, and seamless alignment between human communication, machine data, and service workflows in complex satellite gateway environments.
Plan
The project advances from Kick-Off to Final Review, with core deliverables driving progress: Dev Infrastructure, Storage & Data, Data Architecture, Knowledge Map, AI Framework, GAI Prototype, and Integration.
Progress is regularly reviewed with ESA through the Preliminary Design Review (PDR), Progress Meetings (PM) 1 & 2, Mid-Term Review (MTR), and Final Review (FR). Iterative sprints enable continuous refinement, maintain alignment with objectives, and drive progress against key deliverables. This approach ensures systematic advancement, mitigates risks, and delivers a validated, high-quality AI solution that meets project goals and stakeholder expectations.
Current Status
The project has successfully completed its Kick-Off Meeting (KOM) and has progressed through the PDR, during which the system, software, and data architectures were finalised and validated.
Following PDR, the team is now focused on establishing the cloud infrastructure required for GAI’s environments and pipelines. In parallel, data preparation activities are underway for each initial use case, ensuring that ingested datasets, schemas, and transformation logic are aligned with the approved architectures and ready for integration into the GAI Engine.
The project is now transitioning from design into structured implementation, with foundations being laid for ingestion pipelines, background capabilities, and the first iteration of the GAI App.
Related Links
Companies
