Decentralised professional networking
The goal
A Web3 startup needed to launch a decentralised networking platform to match talent with opportunities at major global crypto conferences.
The paradox
Tight deadlines meet an ambiguous problem space • Fluid project parameters clash with a volatile funding model • Greenfield opportunity risks momentum paralysis
The context
Extreme ambiguity • A fragmented, distributed team • A founder with high vision requiring tactical execution support
My role
I stepped in as Principal Designer to help stabilise product delivery, guide a chaotic pre-release phase, and define a compelling on-chain agentic matchmaking experience with mass-market appeal.
Leading design for an ambitious proof-of-concept Web3 and AI-powered networking platform, I collaborated directly with the founder, CTO, product owner, tech leads, and a compact engineering team in a lean, high-trust environment. I delivered a fully functional proof of concept within four months through fast-paced weekly sprints, combining deep domain research with rapid experimentation.
I pioneered a human-centred design approach tailored to the unique challenges of blockchain UX—simplifying decentralised concepts for mainstream users without compromising technical integrity. I translated complex Web3 interaction patterns into intuitive, inclusive experiences by evolving existing paradigms for accessibility, clarity, and trust. I embedded design throughout the product lifecycle, from early discovery through to front-end implementation.
The challenge
The product MVP initially aimed to solve the "networking problem" at Web3 conferences through decentralised on-chain matchmaking. Yet, there was no clear design pattern to follow. Was it LinkedIn for Web3? A dating app for co-founders? An agent-based assistant?
The founder's vision: An AI agent that autonomously networks on a user's behalf, delivering highly curated, trustworthy, and meaningful matches.
The problem: The team hit a wall of 'analysis inertia' and struggled with unvalidated user journeys.
Technical constraint: The platform had to operate on a specified blockchain (for treasury reasons) but needed to appeal to Web2 users (LinkedIn crowd) without the friction of the closest known equivalent design patterns (crypto wallets).
The strategy
"Walking skeleton" first: I partnered early with the Lead Frontend Engineer to define a basic "walking skeleton" navigation structure—prioritising functional key flows and core navigation—to rapidly expose and test hypotheses and unblock engineering exploration.
The “Walking Skeleton”: Rapid visualisation of potential futures helped unearth key assumptions and drive team consensus.
The "smart profile" concept: I led the design of a conceptual pivot from a standard "bucket" profile (static data) to a dynamic network model—where people derive trust from network validations (who you know) rather than what someone says.
Gamification over corporate: The strategy was to make profile creation feel like an RPG Character Sheet—fun, tactile, and low-stakes to encourage data entry, rather than a dry resume form.
What if? Rapid ideation sketching becomes the impulse for wireframe flows ready for immediate validation testing.
The messy middle
Friction: The visionary cycle
The founder’s high-energy ideation cycles risked derailing development sprints. The result? Mid-sprint pivots left the product team anxious and underperforming.
My pivot: I partnered with the CTO to establish a “Design Shield”—a 1-2-week buffer to interrogate and hone validated concepts before they enter the engineering backlog. By filtering the “signal from the noise”, we ensured the founder felt heard and validated (with rapid prototypes), while the engineering team remained focused on a stable, prioritised roadmap.
Friction: Trust in the unknown
We learned early that high-value industry kingpins (a key target persona) were sceptical of AI matching, uncontrolled exposure, and the risk of inbox bombing.
Our pivot: We moved from a “magic black box” AI to a transparent, plain-language suggestion model, where privacy is paramount. Instead of just serving lists of “potential matches”, the AI-guided platform promised few yet high-quality results, contextualising each with multiple real, tangible, and unique match characteristics.
Instead of “Here’s a profile with a 92% match”, the user sees
Here’s a valuable match because they’re looking for a software engineer like you with blockchain security focus and experience in ERC-721/dApps, and they share your interests in skydiving.
This built relevance and trust by keeping the human-in-the-loop.
Friction: Distributed chaos
The team was global (Melbourne, Sydney, India, Europe, and the US), siloed, and had no practice of context-sharing. Handovers were failing.
The pivot: Together with the lead engineer, I instituted "Baton Pass" rituals and guided the uptake of standardised tools (Linear/Slack/Figma) for documentation and handoff, to ensure the AU team didn't wake up to broken code (or broken commitments) from the EU timezone. The requirement for documentation needed careful balancing with the need for speed.
Less is best: We hypothesised that industry-leading users will connect their existing networks into the platform if they trusted the promise of high-quality over volume.
Stakeholder management & collaboration
Bridging the abstract-concrete gap
The founder and the product owner operated at a high level, while the lead engineer needed immediate, concrete inputs to build the "walking skeleton." I acted as translator between these two worlds. I visualised the visionaries' complex "optimisation problems" into tangible wireframes that the engineer could implement immediately. This broke the "analysis inertia", allowing us to validate high-risk assumptions through code rather than endless surmising.
The outcome
Alpha Launch Readiness: Delivered a functional, testable "Alpha" build in time for the target conferences, giving the on-ground team tools to capture real user data via QR codes and "interest forms."
Stabilised Delivery: Transformed a chaotic, fragmented team into a disciplined delivery unit with clear sprint goals, defined "definitions of done," and a culture of E2E testing.
Strategic Asset: Created a "Smart Profile" architecture that allowed for decentralised, privacy-first networking—a key differentiator in a market weary of LinkedIn's data harvesting.
End-to-end design: I was responsible for the design translation from ideation to hi-fi components. My process was pragmatic, with the underlying principle of just-in-time-design ensuring that outputs matched both the pace of ideation and the requirements of development. Rough, simple, effective deliverables transitioned gracefully to refinement, detail and standardisation for each ‘design chunk’. I used AI-powered UX/UI design platforms (such as UX Pilot) where they made sense, to elevate the design workflow and match quality with efficiency.
When I think about it…
Ambiguity is a feature
In novel Web3/AI spaces, there’s no time to wait for requirements. It’s fruitless to search for best-practice paradigms to emulate—they don’t yet exist. We must take the plunge into ambiguity, “designing ahead,” to create the criteria.
To define tomorrow’s interaction patterns, we can look to the first principles of Design Thinking. A beginner’s mindset, primed with curiosity and a knack for listening to the undercurrents of a user base, will discover new pathways and paradigms faster than a design process of emulation and evolution. The “walking skeleton” approach was more valuable than any high-fidelity mockup because it generated the necessary momentum.
Trust > Tech
The product’s success wasn’t the blockchain tech; it was the UX decision to make privacy, quality, and “volume control” (stopping the spam) the core value proposition.
Newer
Older