jling (@jefftyling) profile picture

co-founder @goldskyio: organically sourced, hand-harvested streaming data pipelines

hip rural location
Joined February 2013
Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)Photo by jling (@jefftyling)
Love tigris - we serve 100k+ reqs/s from our data lake on it. couldn't happen to a better team.
🚀 Dropbox plans to acquire Tigris, and everything you rely on today stays as it is: the product, the API, the pricing, the people. Even more storage, free to move to any cloud. Here’s what it means and what comes next: tigrisdata.com/blog/dropbox-…
1
11
923
🚀 Dropbox plans to acquire Tigris, and everything you rely on today stays as it is: the product, the API, the pricing, the people. Even more storage, free to move to any cloud. Here’s what it means and what comes next: tigrisdata.com/blog/dropbox-…
4
6
6
48
6,235
If you're an open-source user of @ERPC, now you can take advantage of Goldsky's Boost CDN but keep all our own configuration and upstream logic. Boost users end up making up to 90% less RPC calls, with many indexing workloads being 100% served from cache!
New in eRPC: plug in @goldskyio's gRPC cache. Historical blocks, logs, txs and receipts come fast from data Goldsky has already indexed, not from nodes that can return missing or wrong results. 22 chains, one connector: docs.erpc.cloud/config/datab…
1
7
177
jling retweeted
"Stablecoins are becoming the settlement leg." @kevintheli, Founder of @goldskyio, on "Money That Moves Itself" at Ethereum Stablecoin Summit.
5
1
15
593
After experimenting with a lot of dbs for our agent knowledge bases, outside of huge data scenarios, all we need is 1. fast FTS (used as the initial filter before rank) 2. fast count (for basic questions like 'how many blah..') 3. fast ranked relevancy searches the sota changes often but all things lead back to postgres
4
3
311
9ms point lookups without running your own database in every region That's what @goldskyio built with Tigris, @vortexdotdev, and SlateDB Get our side of the story 👇
2
3
9
809
Introducing Feeds: Enriched onchain datasets you can call as an API or stream to a database. Transfers & Balances Feeds are live today. Full historical data, USD values, normalized across EVM and non-EVM, & more soon. Flat price. No minimums. golds.ky/DAPtzjl
5
4
1
52
4,161
Not long ago I'd have called building your own data lake format a waste of an engineering team's time. We built one anyway 🤡 It has been powering all our core products for months, and I'm excited to share how we did this. goldsky.com/blog/prism-geo-r…
1
2
2
15
3,416
for any investors: i'm advising a team working on an machine-code level auto-optimization engine for production ai inference, not a scam. research team includes openCL spec designers, dijkstra prize recipients, etc they have a lead and filling out the round, dm for intro
7
370
day 1 support on all our products, including Boost, for arc! Working with the pre-launch partners, I can already see the energy around Arc and the unique trade-offs the @circle engineers made pay off. As an aside: Huge shoutout to Circle for all their contributions to the erpc open-source repo. It helps make blockchains safer for everyone!
Builders on @arc: hit our CDN instead of your RPC bill. Boost now sits in front of your existing Arc endpoint. Same provider, same keys, reads answered in milliseconds, and users save 45% on average. Free for a limited time. Turn it on under Boost in the dashboard.
7
359
Introducing Boost: a CDN for onchain data that cuts your RPC bill. Your app hits Boost instead of your provider, saving you money. Users save 45% on average (some up to 80%) Bonus: responses as fast as 4ms Works with any RPC provider. Free for orgs for a limited time.
13
13
3
95
10,323
solana dex datasets, with full historical to real-time data with a single pipeline on Turbo!
On Solana, answering a simple question like "Which DEX offers the best SOL/USDC price?" is still difficult. Each DEX speaks a different language, so you can't compare them by default. The US stock market had this problem in the 1970s. 🧵
1
7
713
On Solana, answering a simple question like "Which DEX offers the best SOL/USDC price?" is still difficult. Each DEX speaks a different language, so you can't compare them by default. The US stock market had this problem in the 1970s. 🧵
3
4
2
9
1,639
[Hypercore Data Guide for Analysts and Engineers] 1/N @HyperliquidX consists of two main components: HyperCore and HyperEVM. HyperEVM is not unfamiliar for most data folks — it follows the standard EVM data model: blocks, transactions, traces, and logs. Hypercore powers Hyperliquid's on-chain exchange. It maintains the order book, matching engine, margin system, and clearinghouse state — and exposes a completely different set of datasets. In this thread, we'll discuss what Hypercore data is, where it comes from, and how it's produced.
1
1
3
369
This week, we open sourced Streamling, the engine behind Goldsky's pipelines. It powers data behind the biggest apps in crypto, and built for a world where agents write most software: the runtime guarantees correctness no matter what wrote the code.
1
11
5
49
4,973
Streamling is easy to run in prod. It's a single rust binary following 12 factor principles. Control state can be kept in sqlite or postgres, and it starts up in seconds. The pipelines themselves are stateless. Since our launch, devs deployed 1.3k+ pipelines and jobs.
1
174
We're going to be showing off more of how the design trade-offs we made that allowed us to increase our feature velocity, save 6 figures in cloud bills monthly, and do things we weren't able to do before. One cool thing we wrote about already: goldsky.com/blog/making-hist…
2
139