Data Scientist: turning data into insights and progress, using AI accurately and scalably. Ex-@deliveroo / Ex-Baton (https://xitter.cf/t.co/wyydbAvVuT devco) Founding DS + Head of Data
London, United Kingdom
Joined July 2026
- Tweets66
- Following79
- Followers27
- Likes166
I think the last two months have shown that product & strategy trump everything 💯
unlimited money won’t overcome lack of strategy + bed shitting + thoughtless vibe coded crap 💩
blockworks.com/analytics/sol…
“I fought the markets and the markets won”
Breaking news: The Federal Reserve has raised interest rates by a quarter point to a range of 3.75% to 4%, as widely expected by economists.
Follow live updates: ft.trib.al/PTK3mUo
Wanted to take a look at buyback solana:pumpCmXqMfrsAkQ5r49WcJnRayYRqmXz6ae8H7H9Dfn activity (can only see on chain DEX data from Dune/Allium), which is only ~30-45% of all activity on the token. Interestingly, ~20-40% of all DEX buys is from the two buyback and burn wallets.
Lastly, given the recent rally in solana:pumpCmXqMfrsAkQ5r49WcJnRayYRqmXz6ae8H7H9Dfn it does suggest that CEX trades are probably much more buy-heavy than DEX, which tend to be more sells.
Note any vesting related sells would not have a buy aspect (I excluded transfers)
The codes are all here btw: github.com/odonovse/Fun-Stuf…
I double checked the Dune DEX query on Allium the discrepancy is small ~5%. When I get time I will write the equivalent Python notebooks (or at least Claude will) 🤖
Have first hand experience of this can confirm
The standard explanation for Britain’s declining graduate wage premium is that higher education expanded too fast and too far. But similar expansions elsewhere have not produced the same outcome, writes John Burn-Murdoch. ft.trib.al/RvpW1Vg
I have wanted to understand what exactly is the impact of stablecoins on Solana, especially with the growth of USDC denominated coins and pairing.
At first glance looking at both charts of Stablecoin Supply (green) and Solana price (blue) there is no clear relationship
Worth remembering that this is causal inference, so we are taking correlations and trying to imply causality: could well be wrong. The evidence for a quadratic relationship is clear though (~13% reduction in unexplained variation when included + stat. sig relationship)