Learning Outcomes Addressed: 

• Explain the evolution and business implications of Agentic AI
• Design and build no-code agentic workflows for real business use cases
• Identify and prioritise high-impact Agentic AI opportunities across business functions
• Plan, run, and scale Agentic AI pilots into operational workflows
• Apply governance, risk, and accountability frameworks for AI agents
• Lead the adoption of Human + Agent teams within the organisation
• Demonstrate leadership as an Agentic AI decision-maker

Capstone Overview

This capstone serves as the integrating exercise for the NTU Agentic AI Executive Programme. It requires you to apply everything covered in the programme — agentic mindset, workflow design, no-code prototyping, responsible AI, scaling thinking and leadership framing — to a single realistic business challenge.

Please take a moment to listen to our Professors Alvin NG Theng Haw and Ian Yong Hoe Tan as they explain the significance of this capstone.

Here is the transcript Download transcript of the Capstone Overview video.

The Scenario

The Company

Kanvas & Co. is a Singapore-headquartered D2C home décor and lifestyle brand operating across India, Singapore, the UAE and Australia. The company is seven years old, employs around 140 people and generates approximately ₹180 crore (USD 21 million) in annual revenue.

The brand sells handcrafted planters, wall art, lamps, soft furnishings and candles through its own website, three regional marketplaces (Amazon, Lazada, Noon and Myntra) and twelve experience stores across major APAC cities. It has moved beyond the start-up stage but has not yet reached enterprise scale. Growth remains strong, but operational scaling has become the primary constraint.

Your Role

You are the Head of Growth & Customer Experience, reporting to the Chief Operating Officer. You are responsible for marketing, customer experience and the connection between the two functions. Your team consists of twelve people based across Singapore, Bengaluru and Dubai.

Over the past two years, you have worked to achieve more with limited resources. However, you are reaching the point where strategic improvements alone can no longer create the capacity the business needs.

The Mandate

Last Tuesday, the CEO, Mei-Lin Tan, sent you a one-line email. She wants you to redesign a business workflow from end to end using Agentic AI. She wants a working prototype, a plan for scaling and a pitch she can present to the board. You have ninety days.

“I want one workflow, redesigned from the ground up, with a prototype that runs and a plan that scales. If it works, we tell the board this is how we operate.”

The Workflow Under Redesign

You have selected the New Product Launch-to-Market workflow. This is the process through which a new SKU moves from merchandising sign-off to being live across all channels, with marketing campaigns launched and customer experience teams prepared to support customers.

Today, the process takes between six and eight weeks, involves nine teams and is one of the most frequently criticised cross-functional workflows within the company. It is the process most employees would identify as a major source of delay and inefficiency.

You selected this workflow because it spans both marketing and operations, areas within your remit. It also involves a number of important decisions, including positioning, claims management and channel adaptation, while containing friction that can be measured and analysed.

The Decision Augmentation Hypothesis

This workflow is not underperforming because people are working slowly. It is underperforming because nine teams repeatedly pass the same product brief between them, with each team adding its own interpretation. There is no shared system to identify when claims drift from the original source or when channel-specific requirements are overlooked.

The hypothesis is that Agentic AI can improve the workflow by augmenting three specific decisions: how a product brief is transformed into channel-specific listing content, how that content is checked against brand voice and compliance requirements and how customer experience teams are prepared for launch.

The objective is not to replace the people making these decisions. The objective is to provide them with an agent that creates the first draft and a system that identifies issues that humans may miss.

Capstone Roadmap

The capstone progresses through five structured phases. Each phase builds on the previous one, gradually transforming a business workflow into an AI-enabled operational and strategic solution.

Important Instructions

Please follow the submission process carefully:

• You will receive the Capstone Submission Template during Phase 1 of the capstone.
• The same document must be used throughout the capstone journey.
• For each phase, you will complete the relevant section within the template.
• You do not need to submit the document after every module or phase.
• Instead, the document will progressively evolve as you complete each milestone.
• The final submission will contain all completed sections integrated into one consolidated proposal.
• You will submit one final capstone document during the Capstone week.

Estimated Duration: 20 hours

Note: This is a required assignment and counts towards programme completion.