Define requirements
Review processes, available data and access permissions. Select pilots and establish baseline cost, latency and accuracy.
03 / THE NEXT CHAPTER
PLANNED DEVELOPMENTConnected today.
Intelligent tomorrow.
We’re building on our telecommunications experience to explore what data and AI can make possible for business.
01Data foundations＋
Authorised business data, network metrics and service knowledge, supported by access controls, cleansing and retrieval indexes.
02Models & applications＋
Evaluating suitable model APIs, deployable models and retrieval augmentation for knowledge, speech and operational insights.
03Australian compute＋
Evaluating local GPU inference nodes. Deployment scale and timing depend on application validation, budget and colocation conditions.
04Responsible expansion＋
Starting with requirements and limited pilots, then deployment and expansion guided by validated demand. Monitoring, access controls and human handover are part of the intended operating model.
A direction, thoughtfully developed. This roadmap describes planned capabilities, not currently available GPU capacity or a committed launch timetable.
DEVELOPMENT / THE PATH
Progress with
a purpose.
A proposed development path, subject to assessment and approval. No committed launch timetable is implied.
Run limited pilots
Compare suitable APIs and deployable models. Validate retrieval, speech processing and integrations; measure resources and task costs.
Deploy, then expand
Deploy after budget approval and confirmation of colocation conditions. Expand based on utilisation, demand and cost recovery.
APPLICATIONS / INTENDED VALUE
Better questions.
Practical possibilities.
Expected benefits will be evaluated through actual applications and operational performance.
Customer support
A planned knowledge base and retrieval augmentation could help support staff access product knowledge, procedures and ticket experience.
Business voice
Subject to customer authorisation and applicable processing conditions, explore transcription, call summaries and automatic ticket creation.
Network operations
Combine authorised alerts, performance metrics and ticket data to explore anomaly detection and fault correlation.
VALIDATION / BEFORE EXPANSION
Measure what
makes a difference.
Intended outcomes need evidence. Evaluation will guide the scale and shape of future services.
Quality
Assess answer accuracy, summary quality, false positives and the effectiveness of human handover.
Performance
Measure response times, concurrency, handling time and resource consumption under representative workloads.
Sustainability
Review unit service cost, utilisation and customer demand before expanding infrastructure and applications.