Harness AI Forum (Study Trip)
Architecting the Frontier-Ready Enterprise Agentic Control Plane
Date
2 September 2026, Wednesday
Time
9:30am – 1:00pm
Venue
Garden Studio @ Andaz Singapore, by Hyatt
5 Fraser St, Andaz, Singapore 189354
REGISTER

FEATURED SPEAKERS

Chang Sau Sheong
CHANG SAU SHEONG
Deputy Chief Executive (Products) & Chief Technology Officer, GovTech Singapore.
Previously held senior digital and technology leadership roles at Temasek and SP Group, driving its transformation from a utility company into a regional energy player. 28 years of industry experience, built product engineering teams at PayPal, Yahoo and HP, and authored five programming books.
Prasanth Kumar
PRASANTH KUMAR
Senior Advisor, SPARK (ITGRC); Head Tech Architecture, Engineering & Operations (Former), Monetary Authority of Singapore (MAS)
Joe Chiu
JOE CHIU
Information Technology Consultant, Changi Airport Group
David Chin
DAVID CHIN
Chief Executive Officer, SPARK

The Harness AI Forum is an invitation-only, half-day executive programme for Singapore-based CIOs, CTOs and Technology/Digital CxOs. Designed for peer-level discourse, it convenes senior enterprise leaders to navigate the agentic AI transition — translating frontier capability into boardroom-ready strategy, engineering reality and enterprise governance.

The forum will also include a dedicated pre-departure briefing for delegates joining the SPARK Digital Leaders Summit 2026 in Hangzhou from 13 to 16 September. The briefing will cover the latest programme and site visit updates, travel and logistical arrangements, delegation guidelines, and what participants can expect during the four-day study trip. Delegates will also have the opportunity to meet fellow participants and raise any questions ahead of departure. All confirmed Hangzhou delegates are strongly encouraged to attend.

Architecting the Frontier-Ready Enterprise Agentic Control Plane

As agentic AI moves from prototype to production, enterprise leaders face a converging set of pressures: protecting AI investments from model obsolescence, breaking through decades of IT legacy, dismantling linear workflows for intent-driven service delivery, and governing a growing non-human workforce. This forum arms technology leaders with the strategic and architectural frameworks to build a durable Agentic Control Plane.

KEY DISCUSSION AREAS
01

The Boardroom Blueprint: Capital Efficiency & Future-Proofing

AI investments are being made against a technology landscape that will look materially different in 18 months. How do you build a capital thesis that doesn't become obsolete before it delivers?

Key Discussions:
  • Model depreciation and the economic life of AI assets
  • Build vs. buy vs. orchestrate — who owns the decision?
  • Measuring ROI in a way boards will actually accept
02

The Chief Engineer's Dilemma: Breaking the IT Legacy Barrier

Agentic AI operates at machine speed and assumes clean, real-time data. Most enterprise IT was built for human-speed, batch-processing, and controlled change. Something has to give.

Key Discussions:
  • Engineering around legacy systems with no API or real-time feed
  • Data readiness as the true deployment bottleneck
  • Change management when agents replace the first steps of someone's job
03

Intent-Driven Service Delivery

Digital transformation gave us better portals and faster forms — but the user still navigates a system designed by IT. Intent-driven AI inverts this: the human states what they need, and the system figures out how to deliver it.

Key Discussions:
  • What intent-driven delivery actually looks like in production
  • The digital portal as a legacy interface — what replaces it?
  • Liability and auditability when AI interprets ambiguous intent
04

Non-Human Workforce Governance & Enterprise Liability

An AI agent that can send emails, call APIs and execute transactions is not a tool — it is an actor. Governing a non-human workforce requires controls that don't yet exist in most enterprise risk frameworks.

Key Discussions:
  • Non-human identity (NHI) as an emerging attack surface
  • Risk-tiered compliance at the point of AI execution
  • Defining liability when an autonomous agent causes harm?
REGISTER

For enquiries, please contact:
Gladys Kan | SPARK
E-mail: gladys@wahspark.com

SPARK brings together a community of technology and business leaders from across Asia.