–50 minutes
Registration & Breakfast
–10 minutes
Opening Remarks
–30 minutes
Governing the Agent: Inside the World's First Rulebook
Show moreShow less
- What Singapore's Model AI Governance Framework for Agentic AI actually requires in practice: agent identity, audit trails, and human accountability
- How financial institutions are reconciling MAS expectations with autonomous action in customer-facing systems
- Where Singapore's approach is diverging from the EU and US, and what that means for firms operating across all three
- Building governance that satisfies the regulator without strangling the deployment
–30 minutes
The Cost Curve of Intelligence
Show moreShow less
- A real-world breakdown of what conversational and agentic deployment costs at 2027 model prices, including the line items that never appear on vendor proposals
- Model portfolio strategy: when to use frontier models, when smaller models win, and how the mature teams route between them
- Why headline savings in industry case studies usually reflect partial accounting, and what a full operating model looks like
- Constructing cost cases that survive scrutiny from finance and procurement
–30 minutes
The Machine Customer Arrives
Show moreShow less
- Customers are beginning to send their own AI agents to book, buy, cancel, and complain. What agent-to-agent interaction does to CX strategy
- Authentication and fraud when you can no longer assume a human is on the other side
- The loyalty question: how you differentiate when an agent, not a person, chooses the provider
- What the early data from agentic commerce pilots says about who wins
–30 minutes
Running conversational CX for Asia-Pacific from Singapore
Show moreShow less
- What to centralise, what to localise, and where the regional model breaks
- Language coverage economics across markets from Japan to Indonesia, and the platform consolidation decisions regional HQs are making now
- Data residency and cross-border constraints that shape architecture before any model is chosen
- How regional leaders benchmark market maturity and sequence rollouts
–30 minutes
Voice First
Show moreShow less
- Why voice AI crossed the production threshold, and where latency, turn-taking, and multilingual speech still fail in the wild
- Deploying voice agents into Singapore's four-language, accent-dense environment, on real numbers rather than demo calls
- Where voice agents are replacing IVR at scale, and the customer segments that still refuse to talk to them
- The trust curve: why preference for human agents rises in some segments even as accuracy improves
–30 minutes
When the AI Acts, Not Just Advises
Show moreShow less
- The point in a deployment where conversational AI moves from advising to acting, and how the liability profile shifts the moment it does
- The control architecture senior leaders are now building around action-taking agents, including authorisation, confirmation, audit trails, and reversal pathways
- The cross-jurisdictional regulatory pressure on autonomous action across the EU AI Act, US state-level AI law, and emerging APAC frameworks
- How the most mature programmes are deciding which actions an AI can take unilaterally, which require human confirmation, and which the AI should never own
–1 hour
Networking Lunch
–30 minutes
Panel discussionWho Answers for the Agent?
Show moreShow less
- A senior panel working through accountability when an autonomous agent takes a materially wrong action on the record
- Mapping decision rights before the incident: what the agent owns, what operations owns, what escalates to the board
- Insurance, liability, and contract terms for agentic systems, and where they are still unwritten
- What gets disclosed to customers and regulators, in what sequence, and who decides under pressure
–30 minutes
Fireside chatThe Second System
Show moreShow less
- A candid conversation with a leader who scrapped a first-generation deployment and rebuilt, on what the first system got wrong
- The organisational scar tissue that follows a failed AI programme, and how to spend credibility on a second attempt
- What changed in vendor selection, team design, and executive sponsorship the second time
- One decision they would now make differently, and what changed their view
–30 minutes
What to Bet On in 2027-2028?
Show moreShow less
- A working forecast of which agentic capabilities will commoditise into platform features by 2027-2028 and which will stay genuinely differentiated
- The bets most likely to pay back for regional enterprises operating under the world's most developed AI governance regime
- Where the customer expectation bar is heading across regulated and consumer environments in Asia-Pacific
- Three or four specific bets, with reasoning, that leaders can defend in their next executive review
–30 minutes
Roundtable Discussion 2
–10 minutes
Closing Remarks
–1 hour 30 minutes
Cocktail Reception
No sessions in this track on this day. Choose another day or show all tracks.
Join us in Singapore
Conversational AI Innovation Summit 2027
- CX Value & Economics
- Context & Knowledge
- Agentic AI Orchestration
- Autonomous CX & Agentic Commerce
- Trust, Identity & Governance
- Voice & Multilingual CX