For the past 6 months, I’ve been using my free time to build Yellowcake Analytics, a free resource about the uranium industry. Everyday it ingests new filings and announcements from dozens of miners. If you’re getting up to speed on the uranium trade, it’s a great place to start
22
2
88
9,423
Patrick Scott retweeted
The llamas just shipped a major upgrade to LlamaAI, the only LLM connected directly to DefiLlama data. LlamaAI now powers a “Today in DeFi” update on the DefiLlama homepage. Free users: you now get 5 questions/day for the next week, and new users get 5/day for their first week. A backend upgrade makes LlamaAI more analytical, smarter, and faster. Try it. defillama.com/ai
23
12
69
11,336
Patrick Scott retweeted
Replying to @patfscott
Contractor here: see this all the time. Its the evolution of lead generation services. Spoiler: only hack contractors will work for these outfits.
1
2
574
I used to spend a lot of time making guides for how people could use DefiLlama data. With LlamaAI, that almost feels like a waste, because the easiest way to find and use DeFi data is in our chat interface.
7
1
22
3,528
Apparently the new hustle is dropshipping home services. Guys run ads for a service they don't know how to provide in a market where they don't live, and then hire subcontractors for fulfillment. I doubt many people are making good money with this, but the gurus are printing.
7
22
3,443
Do larger buybacks lead to better token performance? The answer appears to be: no I used DefiLlama's LlamaAI to look at 31 protocols with buybacks over 389 total 30-day periods. Plotting buybacks as a % of mcap at the start of the 30-day interval vs performance over in that interval, there's essentially no correlation (r=-0.07). Running the same analysis over 90-days isn't any better, with a correlation of r=-0.06.
11
7
50
4,464
If you look protocol-by-protocol there's definitely more of a correlation, especially among tokens with the largest buybacks.
1
2
6
598
Recently, we’ve been adding new datasets to @DefiLlama at the fastest rate ever. These have focused on expanding DefiLlama for two different use cases: (1) Discovery of crypto products, and (2) Tracking investments. DefiLlama has been the top place to discover DeFi protocols and chains for years. But we were leaving on the table a huge number of crypto-related products that didn’t fit into either of those categories. Exchanges - We added a dashboard to compare exchanges. This combined existing datasets we had about exchange reserves, liquidity, and markets, with new data about fees and availability. Crypto Cards - The newest dashboard on DefiLlama. This is targeted towards people that are looking for to discover crypto cards, as a user, not an investor. On the other hand, crypto investors (me included) are often now diversifying into other assets. We’ve been turning DefiLlama into a place that investors can go for all types of investments Tokens - Not exactly new, but we added a new dataset focused on tokens (and information related to them) rather than the protocols associated with those tokens as we had historically done. Equities - DefiLlama now tracks thousands of publicly traded companies, with their financials and market caps framed in terms that will be familiar to crypto investors. Pre-IPO Companies - Following our acquisition of Bulletin earlier this year, we received access to a large amount of OTC data on private market valuations. I also took several side quests this year to acquire data for other alternative investments, like trading cards. None of those are live yet. The cool thing is that the products we’ve built for analyzing DeFi can be easily used for these new datasets as well. LlamaAI now answers questions on both crypto and TradFi. Investors will soon be able to track equities, crypto and other alternative investments side-by-side in custom dashboards.
Which crypto cards pay cashback? Which have no FX fees? Which are self-custodial? Our new crypto card dashboard compares 30 cards side-by-side so you don’t have to. defillama.com/crypto-cards
5
26
4,186
Patrick Scott retweeted
palo alto people are building agents for a world where everyone has 47 tasks at once while the average person is so fucking bored they’re doomscrolling during work hours
147
550
14,304
396,758
Patrick Scott retweeted
Today is my 3 yr anniversary of purchasing a construction company as sole owner. The past 3 yrs have been the hardest of my professional life. I'm both proud to have survived so far and embarrassed that I got myself into this situation in the first place. I purchased in 2023 via SBA 7a, underwritten as a $3M rev business. QofE and my own diligence missed two main things: the seller's second business was floating most cashflow issues and their local reputation was unusable. I rebranded to "Cooper" within 95 days on Jan 1, 2024. This industry runs on relationships and reputation. I instantly became an unknown entity and revenue dropped 95% overnight compared to what was underwritten. There are a lot of stories to tell about 2024, but tl;dr I scratched and clawed back to only $1.7M rev by the end of the year. I emptied my savings and depended on hard money to survive. Throughout that year, I made fundamental changes to our sales and operations. We focused on municipal work because I could personally handle the administrative burden and the playing field felt more fair. But our bid -> sale -> collection timeline ballooned to around 9-12 months. Changes wouldn't bear fruit for a long time. In Q2 2025, our 2024 efforts started to take effect. By EOY, we had grown to $3.65M in rev while improving gross margins. However, we were completely cash poor due to growth capital needs and crazy interest payments. We leaned on MCAs to get through because we couldn't compromise all our new client relationships. This year, we professionalized further, increasing both sales and GM. Our annualized revenue is ~$6m and this will be the first year with substantial EBITDA ($1.2M). We took some less horrible debt and I'm working on a larger refi to stabilize a bit. I am still just trying to maintain overall debt coverage. All this time, I haven't paid myself a cent. My team has remained underpaid. Everything has depended on selling the vision to both employees and clients. Selling the future to fund the present. As I reflect on the last 3 years, I feel optimistic about the future. I am a MUCH better operator than in 2023. My team is lean and loyal. And any further growth should be upside I can capture. But if you asked if it was all worth it, I'm not sure how I'd respond. For the past 3 yrs, I've chronically felt nauseous about paying bills, payroll, potential liability/conflict in our work. I've been less present at home. I've been very boring with friends. My health has taken a backseat to the needs of the business. Not to mention my net worth has tanked compared to pre-purchase levels. Our conservative plan is to hit $8M in '27, $12M in '28, then $19M in '29. I know it will continue to be hard, but I'm excited to outgrow the first phase of ownership and work on a fresh set of challenges. Here's to the next three years.
281
54
2,981
1,193,441
Update on my uranium analytics side project: Have had a steady stream of visitors to Yellowcake Analytics recently, despite not promoting it much on here. A few observations: 1. Spot price is my most successful page, both in terms of number of users and engagement time. It also has a very high repeat visit rate. I was worried this would a fully commoditized dataset, but adding different views has attracted a small, but loyal audience. 2. The map of mine and reactor sites has been one of the most-viewed pages, but has had significantly worse engagement time than I expected. 3. Persistent issues with Google indexing have meant that most of the site still isn't visible on search. Instead, growth has been organic and from me conducting direct outreach within the industry. Top priorities right now are getting the site indexed and finding a way to monetize (as much as I enjoy doing this for fun, I'm spending money on APIs and won't be able to do that forever).
8
37
3,149
This crypto rally has had one noticeable absence: Digital Asset Treasury companies DATs have seen less than $2.5B in inflows over the past 4 months. compared to over $5B in an an average month last year.
2
28
4,034
Do buybacks or dividends perform better for DeFi tokens? I ran the numbers with DefiLlama's AI tool. Buybacks were the definitive winner. Out of 27 tokens analyzed, buybacks had a median return of 15.7% YTD and -37.5% over 1 year, while tokens with direct distributions reinvested had a median return of -3.0% YTD and -68.4% over 1 year.
26
19
115
15,063
I added a category of protocols that are earning revenue but retaining all of that revenue for the protocol itself, rather than distributing to token holders. This category actually performed significantly better than tokens that distributed revenue to holders.
1
1
10
1,318
Here's a key insight with this: The disparity in performance between the two cohorts likely has less to do with the mechanism of value accrual than with revenue performance. Median YoY change on revenue for protocols with dividends is -69% vs -41% for protocols with buybacks.
6
662
We've been quietly building a media arm within @DefiLlama. Earlier this year, I set a goal of building distribution beyond our website and main X account. At first it felt like a distraction, but now we're finally seeing results. Nearly 40K subscribers across our two newsletters, breaking down what our data shows. These have already changed how we launch new products. As we scale email acquisition, I expect this to turn into one of our biggest distribution channels. We launched the @llama account, which users can tag to receive a response from LlamaAI. Llama is now tagged in hundreds of replies per month. More recently, we launched the @FullyVested_Pod, a DeFi podcast hosted by me and @DeFiDave. Since launching over the summer, our videos have had over 60K cumulative views, with most of that in the past month. Owning distribution means that we can get our data and products in the hands of more people, without being beholden to the algorithm.
24
4
183
10,391
Never doomscroll.
Community note
The video is an animation by artist R.J. (@RJ16848519) from his Screen Time series, shared here without credit. x.com/rektober/statu… x.com/connoissurreal… x.com/kikanicolela/s…
1
1
26
4,463
Patrick Scott retweeted
You have one single life anon And one single mortal vessel you will live within and navigate that life with (your body, that is)... The quality of that mortal vessel will be a massive, massive variable in the quality of your life And therefore maxxing out that mortal vessel (getting jacked + fit) is the most +EV decision you will ever make
37
8
143
5,922
The era of DeFi value is here. 125 protocols passed on at least $10K to token holders in the past 30 days. 62 have passed on at least $100K.
15
5
80
5,444
Patrick Scott retweeted
TOKENS WITH BUYBACKS, ACCORDING TO @DefiLlama: HYPE, LIT, AERO, PUMP, DRV, KNTQ, SYRUP, ENA, ETHFI, RAY, JTO, RAY, AAVE, LDO, STONK, ASTER, SKY, NEAR, DBR, PONS, PENDLE, CC, CARDS, AAVE, MET, RLB, JTO, META, KMNO Welcome to the next stage of the revenue meta, lads.
WHY: None of these onchain business tokens are securities. We did not know this until now. This clarity makes this massively bullish. We heard about a “DeFi carveout” for cash-flowing assets in the past, but this the nail in the coffin for onchain finance bears. The floodgates are about to open in a massive way. Trillions are coming onchain. Be more bullish.
257
199
1,483
283,029