The Ecosystem OS
I Co-Founded
A multi-chain hackathon platform that turns one-off events into a continuous launchpad for Web3 and AI startups.
"Hackathons drive a meaningful share of every Web3 ecosystem's developer adoption. They are also held together by Discord, Notion, Google Sheets, and goodwill. Fluxor is the operating system that should have existed five years ago."
01 — The Context
Hackathons matter more than the industry admits. Every major Web3 ecosystem — Ethereum, Solana, Polkadot, Base, Cosmos, NEAR — runs builder programs because the conversion rate from "weekend project" to "real protocol" is high enough to justify the spend. The category quietly moves billions in grant capital and produces a meaningful percentage of every chain's flagship apps.
The infrastructure around it is embarrassing. A typical hackathon runs on Discord for community, Notion for the rulebook, Google Sheets for submissions, Devpost for the gallery, Telegram for mentor channels, and a separate spreadsheet for judging that nobody can find on Sunday at 11pm. Sponsors can't measure ROI. Mentors can't scale. Judges drown. Builders bounce between events with no portable reputation.
Ecosystems
BNB Chain, Flow, U2U, and growing
Event Types
Hackathons, ideathons, bounty sprints
Stakeholders
Builder, Mentor, Judge, Sponsor
Thesis
Hackathons are funnels, not events
02 — The Bet
Hackathons are not events.
They are recurring funnels — and they deserve infrastructure that treats them that way.
— 01
Four stakeholders, one shared platform
Builders need a clean submission flow, mentor access, and reputation that travels with them. Mentors need a way to scale past ten conversations. Judges need triage and decision support that respects their time. Sponsors need ROI measurement on prize pools that often exceed $100K per event. Fluxor was designed around all four — four dashboards, one platform, one shared reputation graph.
— 02
AI as a scaling layer, not a feature
The AI layer was placed specifically where humans cannot scale — mentoring at 3am of day two when one mentor covers a hundred teams, and judging 200 submissions in a four-hour window. AI handles the first layer. Humans make the final call. That division of labour is what makes the platform work at real event scale and what turns a founder favour into platform infrastructure.
03 — My Role
Co-Founder & CMO
Brand & Design System
Designed the full product identity — the logo, brand system, modular UI component library, and visual language for the mascot. Every public touchpoint shares one design system I own end to end.
Four Dashboards
Built unified dashboards for sponsors, judges, mentors, and participants — each tuned to the specific job that role does during a live hackathon, not how a CMS organises content.
AI Integration
Integrated the AI assistant agents for live mentoring and judging automation into the product surface. Designed the interface that makes AI actions legible and trustworthy to every role.
User Testing
Ran multiple rounds of user feedback testing during real hackathon events and shipped the scalability fixes that came out of them. Product decisions driven by live failure modes, not projections.
Partnership Pipeline
Own the ecosystem partner pipeline — chain sponsors, VC introductions for graduating builders, and ongoing partner onboarding across Web3 and AI ecosystems.
Tech + Marketing Teams
Run both the tech and marketing teams. Own brand voice across every public touchpoint, from event keynotes to in-product copy and developer community presence.
Market Positioning
Fluxor's positioning, messaging hierarchy, and how it shows up at developer events are mine. Builder-native, not boardroom-native is the single rule.
Agency × Product Thesis
Fluxor is where the agency thesis and the product thesis converge. The hackathon infrastructure feeds agency outcomes. The agency outcomes validate the platform roadmap.
04 — The Build
Four Roles. One Platform.
Every hackathon runs four jobs simultaneously. Fluxor gives each one a dedicated surface built around what that role actually does during a live event.
Builder
Submit, track, and build portable reputation across every event.
- Project submission + milestone tracking
- Mentor session booking and queue
- Real-time leaderboard position
- Personalised bounty recommendations
- On-chain reputation profile
Mentor
Scale your attention across 30 teams — not just the three who found you first.
- Expertise and availability configuration
- AI-routed team assignments by stack
- AI first-layer agent handles common questions
- Human escalation above complexity threshold
- Session logs and team progress tracking
Judge
Make final calls without drowning in 200 submissions first.
- AI pre-scores all submissions on rubric
- Ranked queue with explainable AI scoring
- Flag submissions for committee review
- Override and adjust AI scores freely
- On-chain auditable result record
Sponsor
Measure ROI and make the case for next year's budget.
- Real-time submission analytics
- Builder demographics and engagement data
- Bounty challenge performance metrics
- ROI calculation vs prize pool deployed
- Top-team handoff for VC / grant follow-on
AI at the Points Humans Cannot Scale
AI Mentoring Agents
The AI mentoring layer is a first-line responder, not a chatbot. It understands the hackathon context, the team's submission state, and the bounty they are targeting. It can pair-program through common bugs, explain protocol-specific quirks, and triage to a human mentor when it reaches its limit. One mentor can effectively cover thirty teams instead of three.
Automated Judging
Two hundred submissions, four hours, twelve judges, a complex rubric. Fluxor runs the first pass against the rubric, surfaces the top quartile, and gives judges a structured workspace to make the final calls. Human judges still decide. The platform removes every part that does not need human attention, then records every scoring decision on chain so the result is auditable.
On-Chain Reputation
A builder's track record across Fluxor-powered hackathons lives on chain and is portable. Submissions, prizes won, mentor reviews, judge scoring, and post-event traction all attach to a single builder identity that travels between events, ecosystems, and hiring conversations.
Bounty Recommendations
Personalised bounty surfacing across active and upcoming events. A team that just shipped a ZK proof rollup gets the next zkML bounty surfaced first. Scattered bounty hunting becomes a developer-aligned feed that rewards specialisation.
Brand & Mascot
The green mascot is a deliberate signal that Fluxor is builder-native, not boardroom-native. The brand is playful where it can be and serious where it has to be. One rule: speak to the developer, not to the procurement office.
"Hackathons are not events. They are recurring funnels. The operating system underneath them should treat them that way."
05 — Tech & Stack
06 — The Result
1,500+
Developers activated
200+
Projects pitched
500K+
Impressions generated
10+
Programmes delivered
5+
Ecosystem partners
GOWTFHACK and additional live hackathons running on the platform. AI mentoring agents in production for live event support. Automated judging across multi-chain submission categories. On-chain reputation live for participating builders. Partnership pipeline active across major ecosystem sponsors. The platform is live at fluxor.io.
07 — What It Unlocked
Fluxor is the project that connects the rest of the Luvon Labs thesis. The agency has helped clients win more than $4M in hackathon prize money, and a large part of why that works is because we built the infrastructure those hackathons run on top of. The agency work feeds the product. The product feeds the agency work.
The project proved that the next generation of developer infrastructure is going to be built by people who have run hackathons, not by people who have only sold to them. Fluxor also gave Luvon a permanent seat in the ecosystem conversations that matter. Hackathons are where ecosystems make their developer bets. Owning the platform underneath that activity is a strategic position the agency could not have reached through services alone.
Agency × Product Flywheel
The agency work feeds the platform. The platform feeds the agency outcomes. Fluxor is the single project where both sides of the Luvon Labs thesis converge and reinforce each other in a loop that compounds over time.
Ecosystem Seat at the Table
Hackathons are where ecosystems make their developer bets. Owning the platform underneath that activity is a strategic position the agency could not have reached through services alone. It changed the nature of the partnerships available.
Proof of the AI Thesis
Fluxor proved that AI as a scaling layer for human attention works in production. The mentor coverage model and judging automation that came out of this platform have influenced every subsequent AI product we have designed since.
"Hackathons are not events. They are recurring funnels, and they deserve infrastructure that treats them that way."
"AI mentoring is positioned as a first-line responder, not a chatbot. One mentor can now cover thirty teams instead of three."
"Sponsors come back because we can prove the spend worked. That single capability turned hackathon hosting from a marketing line item into a measurable acquisition channel."
"Builders bounce between events without portable reputation. Fluxor is the layer that finally lets a builder's track record travel with them."
Start a Project
Want to build with intent?
Let's talk about your product — the real constraints, the real goals, and the cleanest path between them.
Start a Conversation