Read the paper, not the tweet
RLNC, mump2p, sidecars, cross-chain bridges. Before I explain anything in a server, I read the source. A moderator who repeats marketing copy gets caught inside a week.
Dhaka
@shariaronchain
I run Discord for protocols where the users hold real money. Right now that is Optimum: 41,420 members, 2,236 of them online.
Five rooms in five years. The newest one has 41,420 people in it.
Community Moderator
The Discord for an MIT-founded data acceleration network, with $7.7B of ETH stake on it.
Community Moderator
Cross-chain DeFi, now a non-custodial DEX with a published CertiK audit.
Discord Manager
Community and Discord for a staking and NFT platform, from early stage.
Listing Partner
The other side of the table: how projects get judged before a community ever sees them.
Community Moderator
First moderation role. DeFi co-financing, built from an empty server.
What I actually do all day, in three moves.
RLNC, mump2p, sidecars, cross-chain bridges. Before I explain anything in a server, I read the source. A moderator who repeats marketing copy gets caught inside a week.
Nobody joins a Discord to learn network coding. They join because a validator earns or it doesn't. Every explainer starts from that sentence and works backwards to the mechanism.
A thread without follow-up is a press release. I answer the replies in the server for the following week, which is where the real questions show up and where the next post comes from.
Three things that happen in every serious protocol server. My moves.
Why this works, and what I am looking for next.
I studied textile engineering, not marketing. That turned out to be useful: textile is a tolerance business. You learn early that a mediocre process run consistently beats a brilliant one run erratically. A Discord server is the same machine, with worse formatting.
In 2023 I went and got the theory to match the practice — Community Manager School: onboarding, health metrics, KPIs, event triggers. Most of it confirmed things I had already learned by breaking them.
The part I am best at is the part most operators avoid: translating something genuinely difficult into language a stranger will read to the end. That is why protocol research teams keep ending up in my DMs.
What I want next is a protocol that takes its community seriously enough to hire one person to own it — not a team of volunteers and a Notion page.
Thirty minutes is usually enough to know if this fits. Pick a slot.