Bringing radical efficiency to complex construction projects

I led research and design for Axiom Labs' 0 ➔ 1 MVP launch, creating an AI-powered solution to streamline construction project reporting and empower project managers with actionable insights for better decision-making.

Innovating office subleasing for the hybrid work era

I led the design and iteration of Halo’s digital platform, aimed to help companies manage and monetise excess office space in the post-pandemic era. By streamlining space planning and end-to-end subleasing support, we created a full-suite solution that simplified a traditionally complex process, enabling businesses to optimise their real estate assets with greater ease and efficiency.

Role

Founding Product Designer

Timeline

Oct 2022 -
Mar 2023

Industry

Prop-tech

Team

me +
4 founders (GM, product, GTM, CX)
1 developer

Overview —

We partnered with a global real estate firm to explore opportunities to modernise commercial office leasing in a post-pandemic, hybrid-work world. In a market where companies were now locked into long-term leases for underutilised space, we saw an urgent need for a smarter, more flexible subleasing solution.

Through rapid iteration and close collaboration with target customers, we developed Halo, a full-suite platform designed to help corporate tenants market, price, and sublease excess office space with greater ease. By combining space planning tools, market intelligence, and operational support, Halo streamlined a traditionally complex process into a clear and actionable experience.

Challenges & Pivots —

The team was originally pursuing a marketplace for premium flexible offices, but conversations with corporate tenants quickly revealed a deeper need: not for finding new and more flexible space, but for support in managing the space they already had and were tied to.

Industry Challenges

Massive vacancies: many companies have way more office space than they need due to remote and hybrid work


Long-term lease commitments: these same companies are locked into 10-20+ year leases


Rigid leasing models: the traditional leasing process is slow, complex and not designed for flexibility or short-term needs


Market stagnation: landlords struggle to fill space, and tenants are burdened with costs and don’t have the tools to sublease effectively

Corporate Tenant Challenges

Lack of expertise: these companies are not landlords, they don't know how to market, price, or manage leasing processes


No dedicated tools: existing real estate platforms are designed for landlords and brokers, not for tenants trying to sublease their space


Operational burdens: managing listings, pricing strategy, tenant fit-outs and negotiations are time-consuming and confusing


Urgency to act: corporate tenants know they need to reduce financial waste quickly but don't know where to start

These challenges revealed a clear opportunity: corporate tenants needed something to help them quickly and confidently take control of their underutilised space. The real opportunity was in enabling both sides of the transaction: helping companies sublease their excess space while connecting them with the right tenants. Our goal was to provide a solution that met them where they were: overwhelmed, under-resourced, and in urgent need of support and clarity. 

Delivering real world impact —

The V1 MVP successfully addressed critical user challenges by prioritising:


  • Automated data collection and streamlined reporting: This significantly reduces manual effort by at least 10 hours per week, freeing project managers and directors to focus on high-value decision-making rather than administrative tasks like reporting

  • Real-time role-specific insights: Tailored and responsive dashboards ensure that site teams and decision-makers can quickly access up-to-date information most relevant to their specific responsibilities


  • AI-powered risk predictions and recommendations: By combining predictive analytics with an AI copilot feature, we empower project teams to identify potential delays and risks, and plan ahead, shifting from reactive problem-solving to a more proactive approach


Prototype testing confirmed that an AI-powered decision support system is a highly attractive and valuable offering for key stakeholders in the construction industry. This was reinforced by direct user feedback, such as:

"This will save people a lot of time and help make more informed decisions."


Project Director, MENA

"This will save people a lot of time and make more informed decisions."



Project Director, MENA

"I've never come across something like this before. It's a game changer, really."


Project Manager, UK

"I've never come across something like this before. It's a game changer."


Project Manager, UK

"It makes my life as a PM much easier. It can facilitate our work on a daily basis."


Project Manager, UK

With concept validation scores ranging from 8.4 to 9.5/10 across features, the product demonstrated clear product-market fit and significant business potential. We received really positive feedback and many requests to collaborate, leading to the team securing their first pilot customer - an opportunity they're leveraging to keep testing and refining the platform in real-world conditions.

Key Design Decisions —

🎯 Role-based insights
We discovered through our research that project directors, site managers, and executives all needed different lenses on the same project data. I designed a customisable KPI Library system that could serve each role's unique needs while maintaining data consistency - an essential requirement when dealing with high-stakes construction decisions.

🤖 Trust-building automation
One of the bigger design challenges was introducing automation to an industry that’s still heavily reliant on Excel and paper reports. I approached this by first automating their most time-consuming tasks, then designing easy and transparent ways to showing the AI's reasoning and manually verify data. This progressive disclosure approach will help build trust and keep users in control of critical decisions that could impact multi-million dollar projects.


🔐 Complex permissions architecture

With the complex ecosystem of players within a given construction project, it was an intricate challenge of designing permissions across that ecosystem. I carefully mapped user journeys and hierarchies, to ensure executives could access high-level insights and confidential information while project managers could work with detailed site information, all within the same platform.


💡 Intuitive yet approachable AI
While AI adoption still varies widely across industries, the construction sector can be particularly hesitant to embrace new technologies. For the AI copilot feature, I wanted to experiment beyond traditional chatbot interfaces while still maintaining familiar interactions. I incorporated example prompts and a prompt library into the experience to ease skepticism and demonstrate the system's potential.

Get instant insights with the AI copilot and prompt library

Create forecasts for project completion and get ahead of delays

Choose which data you see and how you prefer to see it

Get alerted to patterns and verify with remote monitoring

Learnings —

Challenging Initial Assumptions → Pivoting from a top-down, nice-to-have solution ("build it and they will come") to a user-centric platform that tackled real pain points ("meet them where they are") was key to driving customer adoption and building positive sentiment.


Iterate, Test, Iterate Again → Iteration was at the heart of our process—we tested every aspect of the product, from AI feasibility to workflow integrations with dummy data, and refined at every stage. We never sacrificed quality for speed but knew what the right balance was, when.


Building User Trust → Trust is earned, not given. Demonstrating the platform’s value through small wins early on was critical to gaining customer confidence and stakeholder buy-in.