How Artemis built the Bloomberg Terminal for stablecoins and crypto

Meet Artemis Analytics

Artemis Analytics is the analytics platform of choice for institutional investors, researchers, and financial analysts exploring onchain fundamentals. From stablecoin flows to protocol revenues, Artemis transforms raw onchain data into clear, actionable insights trusted by firms like Visa, Sequoia Capital, Circle, Grayscale, VanEck, Solana, and Tether.

Their journey began when founders Jon Ma (CEO) and Anthony Yim (CTO) saw how hedge fund analysts were manually patching together onchain metrics from multiple sources. There was no source of truth, and certainly not one tailored for stablecoins.

From there, Anthony and Jon launched Artemis in 2022 to solve that gap. By standardizing fundamentals across major chains and protocols, they power research on 12,000+ tokens with 1,000+ pre-modeled metrics.

Even better: Artemis meet analysts where they already work. With integrations with tools like Excel, Google Sheets, and Snowflake, teams can analyze crypto markets using the same platforms they already use for equities.

Broken arrows 🏹

Speed to market is everything. But for Artemis, scaling data access across chains started to slow them down. As they expanded to support more networks, streaming reliable data became more complex and increasingly expensive.

They tried multiple indexers and streaming providers. But none of them scaled cleanly. Costs ballooned in unpredictable ways. Their data vendor spreadsheet quickly ballooned into a complex matrix of pricing models, renewal dates, and migration plans.

But operational complexity wasn’t the only issue. Existing providers came with additional engineering problems:

• Slow, unreliable backfills made it difficult to quickly iterate, test, and ship products.
• Rigid delivery models meant no control over schema, structure, or how data was consumed (and forced Artemis to work around provider constraints, with no data ownership or portability).
• Constant schema changes from providers regularly broke pipelines without warning.
• Poor developer experience slowed workflows and added hidden maintenance costs.
• Slow support for adding new chains meant waiting weeks - or even months - for coverage their customers needed.

After trying nearly every option, what finally worked was Goldsky Mirror: a platform that gave Artemis the control, speed, and flexibility they couldn’t get anywhere else.

“Goldsky solved all our data ingestion problems out of the box. It was cheaper, faster, and more reliable than other providers. The developer experience was painless, and made setting up new pipelines a breeze. The turnaround time for adding new chains is best-in-class, and the Goldsky team is super collaborative when prioritizing new chains and features. The team is also super quick to respond to questions and any support inquiries. It’s the default for when we want to support a new chain.” — Alex Kan, Engineering @ Artemis

A more modular (and cheaper) data stack

Mirror scaled better, both technically and economically. It was affordable to add multiple chains without having to make tradeoffs elsewhere. More importantly, Mirror gave Artemis true data ownership and unprecedented deployment speed.

At the core of Artemis’s setup is a decoupled architecture built around Amazon S3 as the canonical landing zone for raw onchain data. Using Goldsky Mirror, onchain data is streamed directly into Artemis’s own S3 buckets, stored as Parquet files for efficient storage.

Because pipeline configs are simple and single-file YAML, they are easily templated and standardized, letting Artemis quickly spin up support for new chains with minimal engineering effort. It’s as easy as copying a config.

Mirror also lets Artemis route data into isolated S3 buckets, supporting access control, cost attribution, and clean separation between chains, products, and use cases.

Once data lands in S3, Artemis loads it into Snowflake where it’s modeled and used to power products including:

• Artemis Terminal – their flagship no-code crypto analytics platform for stablecoin flows, asset fundamentals, and chain + developer activity
• Artemis API – real-time API for hundreds of assets and fundamental metrics
• Sheets – native Excel and Google Sheets plugins for crypto data
• Datashare – direct access to the most comprehensive crypto data warehouse

“Once I get the thing going it literally never dies”

Goldsky Mirror gave Artemis a modular, scalable foundation for streaming onchain data - one that aligned with how they actually wanted to build. Unlike other providers that dictated format, timing, and schema, Mirror offered full control over delivery and architecture.

This decoupled setup delivered six critical advantages for Artemis:

1. Cost efficiency: Writing directly to S3 avoids the need to keep a Snowflake warehouse always on, reducing compute costs significantly. With Parquet files staged in S3, Artemis only ingests what’s needed, when it’s needed.
2. Granular pipeline control: Artemis defines their pipelines, specifying what data to pull, how it’s structured, and where it goes. They no longer have to worry about vendor-driven schema changes or brittle delivery formats.
3. Modular architecture: Because ingestion, modeling, and analytics are decoupled, Artemis can swap out any part of their downstream stack without rewriting pipelines, reingesting historical data, or losing access to data.
4. Real data ownership and portability: The raw data lives in Artemis’s own S3 buckets, not in someone else’s data shares. This means long-term archiving, reprocessing, and audibility are built-in. And if they ever move off Snowflake, their data goes with them.
5. Rapid, scalable deployment: Templated pipelines mean Artemis can rapidly spin up support for new chains in minutes with copy-and-paste deployment and minimal engineering effort.
6. Reliable and resilient delivery: S3 writes are append-only, making it easy to dedupe downstream if schemas evolve. With at-least-once guarantees, Mirror ensures data doesn’t get dropped, even if a downstream system is temporarily unavailable. Failures are automatically retried with exponential backoff, keeping ingestion dependable without manual intervention.

Results: “We're chasing a lot more Goldsky data now.”

• $200K+ and countless hours saved: Artemis saves over $200K annually by skipping infra and licensing costs - plus 100s of engineering hours.
• 99.99%+ uptime: “The only time a pipeline has failed was when we pushed a bad config.”
• Rapid chain support + backfills: New chains get implemented in under a day, and historical backfills take hours instead of weeks to months.
• Minimal engineering lift: Pipelines are templated and self-serve, requiring just a few lines of config to launch and scale.
• Reduced vendor risk: True data ownership eliminates surprise price hikes and platform lock-in.
• Real-time ready: The streaming architecture is ready to support Artemis’s real-time products as they expand their offerings.

With Goldsky, Artemis dramatically accelerates research and decision-making for both technical and non-technical users, making it a go-to solution for anyone seeking clarity and depth in digital asset analytics.

Unblock your data team and ship faster

👋🏻 Hey, we’re Goldsky. We help teams like Artemis Analytics access blockchain data at scale without sacrificing speed, cost, or control. Whether you’re building analytics products, modeling fundamentals, or expanding to dozens of new chains, we give you full data ownership and flexible delivery without the typical vendor overhead.

Sign up for free or instantly schedule a call to see how we can help you ship faster. No credit cards or forms to fill.

---

The Artemis and Goldsky founders (missing: Jeff)