What I Did

Led the enterprise rollout of Zoom's generative AI assistant across 20, 000+ employees at Macquarie Group. This was a high-stakes deployment requiring coordination across technical teams, business divisions, compliance, and change management.

The Challenge

Introducing generative AI in a regulated financial institution meant navigating strict security requirements, data governance policies, and organisational change at scale. Employees needed to trust the technology while understanding how to integrate it into daily workflows.

My Role & Contributions

• Technical Discovery: Conducted architecture reviews to understand Zero Data Retention (ZDR) implementation, encryption controls, and integration with enterprise systems
• Deployment Strategy: Developed phased rollout approach across eight business divisions, balancing technical readiness with organisational change capacity
• Stakeholder Alignment: Partnered with division leaders, security teams, and risk stakeholders to obtain approvals and coordinate implementation
• Change Management: Created adoption strategy, communication plans, and enablement materials to drive employee engagement
• Governance & Compliance: Developed enterprise change documentation addressing legal, risk, and regulatory requirements

Key Capabilities Delivered

Outcome

Successfully deployed AI capabilities to 20 000+ employees while maintaining enterprise security standards. Reduced administrative overhead in meetings and enabled employees to focus on higher-value work.

What I Did

Designed the product strategy and use case for a RAG-powered AI chatbot to revolutionise internal knowledge access. This was a hackathon concept exploring how retrieval-augmented generation could transform scattered documentation into an intelligent, conversational assistant.

The Problem

Employees waste hours navigating fragmented knowledge repositories; SharePoint sites, Confluence pages, PDF manuals, internal wikis. Finding the right answer often requires knowing where to look, which defeats the purpose of having documentation in the first place.

My Role & Approach

• Problem Definition: Identified the knowledge discovery gap through user interviews and pain point analysis
• Use Case Design: Defined target scenarios where AI assistance would deliver maximum value (troubleshooting, policy questions, process guidance)
• Solution Architecture: Proposed RAG approach using vector embeddings for semantic search and LLMs for natural language generation
• Adoption Strategy: Developed rollout plan addressing user trust, source transparency, and integration with existing workflows
• Success Metrics: Defined measurable outcomes (time-to-answer reduction, support ticket deflection, user satisfaction)

Proposed Solution

Technical Approach

The solution would use vector embeddings to convert documentation into searchable semantic representations, retrieve relevant context, and generate answers using LLMs while always citing sources. This "retrieval-first" approach ensures accuracy and builds user trust compared to pure generation.

Value Proposition

Reduce time-to-answer from minutes to seconds. Enable employees to ask questions naturally rather than navigate complex information architecture. Improve knowledge accessibility while maintaining accuracy through source grounding.

What I Built

A gamified productivity app currently in pilot deployment (live TestFlight beta). OnTrack transforms procrastination into progress through financial accountability, combining behavioural psychology with mobile-first product design over 1+ year of development.

The Core Problem

People don't need another to-do list, they need a reason to actually complete it. Traditional productivity apps focus on organisation, not motivation. Users plan endlessly but struggle with execution. The gap isn't knowledge, it's accountability.

The Solution: The Commitment Jar

Users stake real money on their goals. Complete your task, keep your cash. Miss the deadline, and it goes into the "accountability jar." This creates tangible consequences for inaction, turning vague intentions into binding commitments.

Key Features

Product Development Journey

• User Research: Conducted interviews with students and young professionals to identify accountability gaps in existing tools
• Design & Prototyping: Designed full user flows in Figma covering onboarding, task creation, jar mechanics, and progress tracking
• Usability Testing: Ran moderated tests with 8 participants, identified pain points (financial penalties felt harsh), and iterated with flexible commitment amounts
• Current Status: Live TestFlight beta in pilot deployment phase, gathering real-world usage data

Design Principles

• 1. Motivate: External accountability mechanisms encourage completion
• 2. Gamify: Transform productivity into an engaging experience with streaks and rewards
• 3. Thrive: Build sustainable habits through financial discipline and progress visualisation

Target Metrics

• 70% user activation (complete first goal)
• 50% weekly retention after 30 days
• 30% increase in task completion rate vs baseline
• Average 5 tasks completed per week

What Makes This Different

OnTrack doesn't just organise tasks, it addresses the behavioral psychology behind why people don't complete them. By introducing real financial stakes, social visibility, and gamification mechanics, it creates the external pressure that internal motivation often lacks.