How Components Uses CamelAI to Power Data-Driven Storytelling

About Components

Founded by Andrew Thompson, Components is a media and analytics company on a mission to make value intelligible.

Unlike traditional news or analytics outlets that focus on surface-level price movements or trends, Components explores the “why” behind human and market behavior — uncovering the complex relationships that determine what makes something valuable.

Over six years, the company has evolved from Andrew’s personal publishing project into a full-scale media and data operation powered by a deep technical backbone. Now, Components publishes regular analyses across industries, and its growing data infrastructure serves as the foundation for an analytics platform that can one day power insights for other organizations too.

The Challenge: Scaling Editorial Data Workflows

When Components began, Andrew handled every step — data engineering, analysis, and writing — alone.

“Usually you have people who can write, people who can analyze data, and people who can build data pipelines. Finding someone who can do all three is nearly impossible,” Andrew said.

As the company transitioned from publishing once a year to weekly or daily releases, Andrew faced a key question: How could the rest of his team replicate his end-to-end analytical process without needing to become data engineers?

That question led him to CamelAI.

“Rather than having me do the analysis or another person do it and then hand it off to a writer, we realized that if we used something like CamelAI, anyone could take on that full workflow — and it’s worked out really well.”

Why Components Chose CamelAI

For Components, understanding “everything at once” meant working across hundreds of interconnected tables spanning multiple industries and data domains.

Manual SQL querying simply wasn’t scalable.

“Traversing those tables manually was miserable. Nobody enjoys it,” Andrew explained. “It’s the kind of thing a lot of people have been waiting for AI to replace.”

CamelAI’s text-to-SQL interface gave the Components team a faster, more intuitive way to explore their database — turning questions into structured queries instantly and correctly.

“We can iterate through exploratory questions 20x faster. It’s been a modal shift for us.”

Implementation & Everyday Use

Setup was frictionless.

“It was really easy. There were no bugs, no errors — it just worked and created a direct line to our Postgres database,” said Andrew.

Once connected, the team started using CamelAI for real-time exploration and research, including their debut analyses The Desire Distribution, a deep dive on Substack’s revenue curve, the end of the winner-take-all media markets, and the cleansing power of money.

One notable example: a comprehensive analysis of Substack’s entire business ecosystem.

“We had structured data on individual Substack publications and models I’d built myself. CamelAI wrote the tricky query to aggregate everything and turn it into a histogram — that would’ve been painful to do manually.”

From there, CamelAI became a daily part of the workflow, powering cross-domain analysis across sources like the St. Louis Federal Reserve’s commodity and housing datasets — automatically handling joins, aggregations, and exploratory questions.

“The ability to easily do joins means we just do more of them. It prevents you from having to remember exactly what’s in your data all the time.”

Trust, Accuracy, and Validation

Despite the power of automation, Components still applies a disciplined “trust but verify” process.

“Analysis should surprise you only within reason. If something looks totally wrong, we check it — just like we would in Jupyter or RStudio.”

Andrew’s team reads the generated SQL when needed, validating logic and ensuring alignment with their schema and migrations.

“Nine times out of ten, when something’s off, it’s because of something in our pipeline — not Camel.”

Writers who use CamelAI have their outputs reviewed by a technically trained team member for additional assurance.

“It’s been remarkably consistent. AI’s just really good at writing SQL.”

Comparing Alternatives

Before adopting CamelAI, Components built a terminal-based text-to-SQL prototype in-house — an early attempt to make querying more intuitive.

“Getting the first 60–70% is easy; it’s the remaining 30% that makes all the difference,” Andrew said. “Compared to our homegrown tool, Camel is vastly superior — and it just works.”

Andrew first discovered CamelAI after seeing its Hacker News launch and was drawn to its reliability and product philosophy.

“We liked the team, it worked well, and it’s been really helpful. So we just stuck with Camel.”

Results & Value

• 20x faster analysis throughput
• Democratized access to complex datasets
• Trusted, readable SQL generation
• Smooth integration with existing Postgres infrastructure
• Higher team engagement across editorial and technical roles

Components can now move from idea to publishable insight in days — not months.

Spotlight: Components × CamelAI

Closing Thoughts

Components exemplifies a new kind of hybrid organization — equal parts media, analytics, and data infrastructure.

With CamelAI, they’ve scaled a uniquely human editorial process into one any writer or analyst can perform, combining storytelling with deep quantitative insight.

“We’ve been able to maintain the depth of our work while making it accessible to more people,” said Andrew. “That’s the real value.”

As Components continues expanding its data-driven journalism, CamelAI remains the backbone of their analytical workflow — powering the stories that make value visible.

👉 Want to learn how CamelAI can help your team uncover insights hidden in your data? Try it free or book a demo.