Analytics · SQL · AI workflow
On-chain data and AI-assisted content
Built the public on-chain dashboard the community kept asking for, writing the SQL myself, and an AI research desk that keeps my own X feed current without ever posting for me.
- 22B → 19.6B
- SWEAT supply tracked, burns included
- Hourly
- Dashboard refresh
- 0
- Posts or replies the AI sends itself
At a glance
- The problem
- The community wanted SWEAT's on-chain numbers, and there was no public source for them.
- What I did
- Built a public Dune dashboard, writing the SQL myself, and an AI desk that researches and drafts but never posts.
- The outcome
- A shared source of truth that backed up our burn and deflation claims, and a steady personal feed with me still in charge of every word.
The receipts
The full story: context, approach, examples and results.
Context
Community members kept asking for SWEAT's on-chain numbers: supply, burns, staking and activity. There was no public place to see them, so our claims about burns and deflation rested on our word alone. Separately, keeping an X feed timely and well researched meant hours of reading every day.
The challenge
Give the community numbers it could check for itself, and keep content current without handing my voice to a machine.
What I did
I built the SWEAT Economy On-Chain Overview on Dune. It was a new platform for me and a fresh challenge, and a chance to sharpen my SQL. I wrote the queries myself, with AI as a supporting tool.
The dashboard covers supply and emissions, monthly burns, daily on-chain activity, staking in Growth Jars, and how many steps it takes to mint one SWEAT. It refreshes hourly, and a methodology section explains how every number is worked out, including where on-chain data alone falls short.
I shared it in our Five in Ten Q&A sessions with the CEO and directly in community channels. Community members used it themselves, and the community team used it to back up our claims.
For content, I built Spear Desk, a daily research-and-draft pipeline for my own X account. Each morning it gathers what moved in crypto, NEAR and marketing over the last 24 hours, writes a short brief with sources, and grades every topic out of 25 for relevance, timeliness, voice, engagement and safety.
Topics that score 18 or more are drafted in my voice with a suggested time to post. It also flags posts worth liking, reposting or quoting, and drafts reply options for the accounts I follow most closely.
It never engages directly, and it has no access to. It researches, compiles and drafts, and I edit, decide and post everything myself. A "never" list keeps it away from price predictions, unsourced numbers and knocking other projects.
The dashboard
The supply section: total supply over time with burns as monthly step-downs, alongside daily mints and monthly burns.

01
Supply and burns
From 22B at launch to about 19.6B, and deflationary since July 2023 thanks to monthly burns.
02
On-chain activity
Daily transfers, new wallets, and an hour-by-hour view of when people move.
03
Growth Jars
About 1.81B SWEAT locked in staking, 9.2% of supply, read directly from the contracts.
04
Emission difficulty
How many verified steps it takes to mint one SWEAT: about 8,990, and rising over time.
Results
The community got a public source for SWEAT's on-chain numbers it could check for itself, and our burn and deflation claims came with receipts anyone could open. The dashboard is no longer maintained following Sweat Economy's closure.
Spear Desk keeps my own feed regular and engaging for my followers, and keeps me up to date with what's happening across crypto and marketing. The same research-to-draft approach supported the three X accounts I ran.
Show the numbers, don't just state them. And let AI do the reading, never the talking.
See it live
- Dune
- SQL
- AI
- Research