Melbourne AI Engineering & Infrastructure Summit 2025

In the rush to implement Large Language Models, many organizations are overlooking the strategic value of traditional machine learning approaches. Through real-world examples and practical frameworks, this talk challenges the "LLM everywhere" mindset and demonstrates how a hybrid approach combining targeted traditional ML with LLMs can create more reliable, cost-effective, and safer AI systems.

Discover a proven roadmap for building and deploying enterprise AI solutions that deliver real business value. Explore the essential steps, from strategic planning and resource alignment to streamlined development workflows, that ensure future growth and adaptability.

‍

Explore how technology leadership, data expertise, and strategic vision converge to drive successful AI programs while uncovering collaboration frameworks, shared responsibilities, and strategies to ensure impactful, long-lasting results.

• How can technology, data, and business leaders align on clear roles and responsibilities to ensure cohesive AI program delivery?
• What steps must be taken to prepare data at scale, and how do teams maintain quality while balancing security and governance?
• Which business outcomes should guide AI investments, and how can leaders measure and communicate the value of these initiatives across the organisation?

‍

In this innovative session, attendees will be faced with a series of scenarios that they may face in their roles. Attendees will discuss the possible courses of action with their peers to consider the ramifications of each option before logging their own course of action. 

Results will be tallied and analysed by our session facilitator and results will impact the way the group moves through the activity.

Will we collectively choose the right course of action?

‍

An infrastructure director recounts how they tackled ballooning expenses for GPU-intensive workloads without compromising performance. Discover how resource monitoring, budget forecasting, and flexible architectures can deliver cost-effective AI solutions.

• Monitoring resource consumption and pinpointing cost hotspots
• Leveraging scaling strategies and spot instances to optimise expenses
• Balancing compute performance with financial accountability

‍

A leading AI engineer shares how they consolidated disparate data pipelines into a unified platform, reducing complexity and improving model accuracy. Attendees will learn actionable strategies for orchestrating data flows and ensuring data quality.

• Centralising siloed data sources to streamline AI workflows
• Automating data ingestion and transformation for consistent inputs
• Enhancing collaboration between data and engineering teams

‍

Explore the technical intricacies of designing, deploying, and scaling AI infrastructure. Delve into the tools, frameworks, and architectures that power high-performance AI solutions, and learn how to balance agility, security, and cost-efficiency.

• How do teams architect resilient, high-performance computing environments to support AI workloads at scale?
• How can teams ensure real-time, high-volume data flow for AI?
• Which pipelines streamline model development, deployment, and continuous monitoring?

‍

Roundtable topics to be shared with registered attendees for their selection

AI is no longer confined to proof-of-concepts; enterprises are driving it into production. This keynote dives into operationalising AI at scale, including governance, MLOps best practices, and performance optimisation.

• Instituting governance and compliance for AI applications
• Implementing MLOps pipelines for continuous deployment
• Optimising performance to meet real-time business demands

‍

In this interactive session, participants will explore and debate five hot-button issues shaping AI’s future. Expect divergent views and lively discussion on how these trends could redefine both engineering practices and business outcomes.

• Cloud vs. On-Prem High-Performance Computing – Balancing elasticity, control, and cost
• AutoML Tools – Do they empower teams or oversimplify complex engineering challenges?
• Ethical AI vs. Speed to Market – Should organisations slow innovation to ensure responsible development?
• Edge AI vs. Centralised Processing – Is pushing more AI to the edge truly efficient or overly complex?
• Low-Code/No-Code AI – Does democratising AI risk quality and governance, or is it the key to widespread adoption?

‍