Kanda Colab

A studio for the ambiguous middle

Kanda Colab was the product studio I founded and ran in Sydney, working directly with founders and small teams across digital products, services and internal operations.

I built the studio for a particular kind of brief: the problem was real, the ambition was clear, but the route between them was not. A client might arrive with an early product concept, a fragmented customer journey, two business systems that refused to cooperate or a valuable manual process that could no longer scale. I would start by finding the shape of the problem before deciding what needed to be designed or built.

Kanda Colab worked as an embedded product function rather than a conventional production partner. Strategy, experience design and engineering could happen in the same conversation, shortening the distance between a decision and something a customer or team could actually use.

Embedded, not adjacent

The operating model kept me close to the work. I did not disappear with a requirements document and return at the end. I joined the client's decision-making process: clarifying the outcome, mapping the essential journey, making an early version tangible and refining it with the people who would ultimately own it.

That closeness mattered most while the brief was still moving. Early-stage teams rarely need every unknown turned into a six-month scope. They need the next important uncertainty resolved. Sometimes that means a prototype that makes a product idea testable. Sometimes it means a complete first release. In other cases, the most useful intervention is quieter: removing an operational bottleneck or making an existing service easier to run.

Kanda Colab's role was to identify that point of leverage and carry the work through, without forcing every engagement into the same format.

Giving product ideas a usable shape

Several engagements began before there was a settled product.

For Dime, a social basketball concept, stakeholder and user interviews were translated into the journeys for finding games, joining players and tracking performance. More than 100 screens and a connected prototype turned a broad idea into a product that could be explored, discussed and tested as a whole.

For ezybalance, the challenge was to make tax feel less intimidating for Australian tradespeople. The response combined identity and product design: a calm visual language, a clearer hierarchy around the tasks people needed most often, and a design system capable of holding together close to 200 screens. Brand and interface were developed as one experience rather than separate exercises.

Octadoc required a different kind of focus. An earlier design direction for a template and document builder was developed into an interactive product prototype over a concentrated two-week engagement. The scope was deliberately narrowed around the essential behaviour—creating a guided question flow, organising personal and shared templates, and showing how the finished document would work for its audience. The prototype gave the team a concrete way to evaluate the concept before committing to a larger build.

The screen counts show the depth of those projects, but they were not the point. What clients gained was shared clarity: a product direction that stakeholders could see, use and make better decisions around.

Connecting the parts of a business

Other engagements focused on the operational gaps customers never see but teams feel every day.

Speak More Clearly ran its customer relationships and course delivery across separate platforms. Subscription changes, cancellations and access updates created repeated administrative work and room for inconsistency. I designed the connection between those systems so customer events could flow through to the right course actions, while leaving the client with a clear operating process for maintenance and access changes. The result was less reconciliation, fewer manual steps and a service that behaved like one business instead of two disconnected tools.

For Adbox, a content-led website had to support an actual membership proposition. Editorial publishing, package selection, discounts and recurring billing needed to feel like one customer journey. The work balanced the flexibility of a marketing site with the precision required around payments, giving the team control over its content without making checkout feel like a separate product.

An independent portfolio engagement followed the same approach to ownership. I structured projects, awards and profile content so the client could manage the body of work without returning to a developer for routine updates. A good handover meant leaving them with a system they could continue to use confidently, not just a static finished website.

Making AI useful beyond the demonstration

The studio's later work moved further into applied AI and automation, but I kept the same standard: the result had to survive contact with a real workflow.

One engagement transformed long-form documents into produced podcast episodes. The visible output was audio, but the real design challenge sat across the entire process—organising source material, shaping it for listening, generating narration, assembling sections and delivering a publishable episode consistently.

I designed the system as a repeatable production line, not a one-off experiment. I used existing automation tools where they were strong and added custom processing where the quality of the final programme needed more control. The engagement also ended with operating documentation so the client could run the workflow without permanent technical support.

That distinction became central to how I approached AI work at Kanda Colab. A compelling output proves that something is possible. A dependable process, understood and owned by the client, is what makes it useful.

Decisions that shaped the practice

Across a varied portfolio, I kept returning to the same set of choices.

Start with the business constraint, not a preferred service or tool. Make the most important part of the experience tangible early. Build enough of the surrounding operation that the solution can function in the real world. And design the handover at the same time as the product, so ownership does not disappear when the engagement ends.

Those principles allowed the studio to move between brand, product, commerce, integration and automation without losing focus. The common thread was not a particular medium. It was turning an unclear, cross-functional problem into something coherent that worked in practice.

The outcome

Through Kanda Colab, I delivered client work across product strategy, identity, experience design, software and automation, from early concepts to systems used in day-to-day operations. It gave me a model for working at my best: close to the customer, accountable for the outcome and able to move across disciplines when the problem required it.

The studio was wound down at the end of 2025. Its operating philosophy continues to shape my work—embed deeply, reduce ambiguity quickly and leave behind something useful, understandable and ready to move forward without me.