Fern generates SDKs and API documentation that companies are proud to share with their customers. Since launching in 2023, we’ve been trusted by over 150 customers, including industry leaders like Square, Webflow, ElevenLabs, LaunchDarkly, and Intercom, to power their developer experience (DX) and agent experience (AX).

Fern’s team of under 30 people works out of headquarters in Brooklyn, New York. The company launched during the Y Combinator accelerator program in early 2023.
 

Frictionless onboarding

Fern started working with Algolia in April 2021 with an open-source documentation tool called Docusaurus. The company needed a search tool for documentation and DocSearch by Algolia was the recommended path.

When Fern started growing in 2023 with it’s own API Docs product, it stuck with Algolia. Not only was it Y Combinator affiliated, but Algolia’s startup program made it easy to integrate and provided a free year of service. “The startup program made it a no-brainer to start building on Algolia from day one.” recalls Danny Sheridan, CEO of Fern. It’s what Fern continues to use exclusively.

Fern built its static sites generator, which is server-side rendered, entirely from scratch. It runs a multi-tenant Next.js app hosted on Vercel to power hundreds of customers' documentation websites. 

It took Fern less than a week to design, implement, then iterate on the Algolia record data structure. After the initial setup, there was no maintenance required. Algolia simply operated the way Fern wanted it to.

So good, we kept it

That early experience was impactful for Fern and Algolia has become a core part of the product. “Algolia is too far in for us to rip, which is a great place to be. It just performs.”

The company thrives on having a product that its customers recommend to their friends. Its goal is to keep growing at an aggressive quarterly rate in revenue as it takes on new business.

The reality is that DocSearch is table stakes for Fern’s product offering. Every customer buying a documentation tool expects search to come wired up and working well. Uptime is vital and Fern’s experience with Algolia has always been perfect. 

The test came during a recent period of rapid customer growth, where Fern tripled its customer base within a year. Customer expectations remained the same, but now Fern needed a reliable search infrastructure to support a 30% month-over-month increase in searches and avoid issues related to downtime.

As Fern’s user base scaled, it was challenging to predict future needs, especially around the number of searches, records, and events.

Algolia continued to hit the mark with search uptime. “Algolia was the right scalable solution to do just that,” says Danny. “I’m never worried about hitting a limit.”

Features that fit

User permissions and restrictions are critical for Fern’s business case. One of the Algolia features Fern uses most is the ability to generate a secured API key which restricts user access to a subset of an index. 

Fern uses the search api key provided for its Algolia index and crypto to generate an HMAC using the original search api key as a signing secret. The logic Fern devised lets the company generate the API key closest to the user viewing the docs. That method runs seamlessly, without needing a network call to Algolia’s servers and responding in under 200 ms anywhere in the world.

Fern also benefits from Algolia’s headless structure, using react components and css files to bootstrap its own DocsSearch experience. The reset.css file lets Fern fully control its own design, stripping out initial styles and implementing its own visual framework and keybindings. For example, Algolia’s snippeting feature — which truncates and highlights relevant pieces of a user’s search query – is built in, but Fern leverages it using its own highlighting styles.

With Algolia as an API-layer, Fern has the foundation and the flexibility to create new API routes, including facet filters and AI chat suggestions. Fern has plans to implement a UX improvement to 404 pages, using the URL slug to determine the most likely page for a redirect.

Paying for scale

After the Startup Program, Fern adopted the pay-as-you-go model. As the company evolved, it approached the scaling process with an MVP mindset, testing different plans and considering pay-as-you-go versus Algolia’s annual payment plan.

The big challenge was always estimating future search volume and records. Different timeframes yielded drastically different records data, impacting decision-making. “We have a rapidly growing customer base,” explains Danny. “Using trailing 30 days versus 90 days versus 365 days gives us completely different numbers. That’s the tough part—figuring out how to predict the number of searches and records we need in the next year.”
 

Eventually, as Fern took on more customers, it was compelled to shift gears. The fact that Algolia only charges by the number of records and searches – factors that Fern can control – made it easier to switch. Fern was able to perform on-demand revalidation and reindexing without worrying about cost. Fern eventually found the scale plan to be more economically beneficial, enabling Fern to handle searches and records more effectively while growing rapidly.

That decision, despite initial uncertainty, paid off. Fern is able to avoid potential infrastructure bottlenecks and provide an uninterrupted experience to a growing number of customers. In terms of competitive advantage, it’s an enviable and economical position for Fern to be in.

A productive partnership

Algolia has been an essential partner in scaling Fern’s infrastructure. The freedom and flexibility to expand seamlessly has been crucial for the open source developer in meeting its growth objectives. The company can rely on Algolia to handle more records and searches while it focuses on product. Fern engineers spend time developing and finessing features that serve customers better, instead of grappling with search or uptime. 

The partnership between Fern and Algolia is consistently collaborative and transparent. Developers partnering with developers makes it productive, too, and there’s a willingness on both sides to engage deeply to improve the developer experience.

“A neat outcome from our partnership would be working with the Algolia team to level up your developer experience, especially around Docs and SDKs,” says Danny.

When companies learn from each other, it’s a win-win.