We are super excited to announce the launch of Appbase.io 2.0, the open core search stack for building modern apps. Over the past four years of building appbase.io, we have realized that it takes so much more than a search engine to build and run search apps at scale.
What’s a search stack and why build it?
There are so many things that today’s state-of-the-art search engines like ElasticSearch or Solr don’t do out of the box, yet are needed before you can deploy your search app to production:
- The development experience — There’s a fair learning curve to tuning a search engine to your needs. Data ingestion from your favorite DB (think Postgres / MySQL / MongoDB) isn’t straightforward. Unlike your SQL database, ElasticSearch’s mappings (aka schema) are immutable. Once set, you can’t edit or delete them. This can lead to a frustrating development lifecycle while you are still figuring how everything works.
One of the first tools we open-sourced: Dejavu addresses some of the major pain points around optimizing the developer experience. It’s actively maintained and has been downloaded over 500,000 times.
The hosted appbase.io service goes further: by offering a CLI for importing data from your favorite source, a GUI for editing mappings and a visual search UI/Ux building experience. Read more here.
- Building the right Search UI/UX — This is another area where you are on your own. Translating a search engine’s query DSL to search UI views that can wow your users is a big leap.
With ReactiveSearch, we open-sourced a set of UI libraries for React, React Native and Vue.JS to address this very issue. You can create amazing UIs with a declarative syntax, using query DSL is optional. In just a short one year, it has become the most popular search UI kit, powering hundreds of production search apps.
With appbase.io, you additionally get production support from the engineers that built ReactiveSearch.
- Securing your search app — Securing your search app / engine isn’t trivial. Even the best hosted search services don’t offer anything beyond a basic authentication support.
We didn’t realize how big of a missing piece this was initially. We only learnt about this over time and in a hard way from our customers.
With appbase.io, you can define fine-grained access rules, whitelist IPs and HTTP Referers, programmatically create / update / delete API keys or set a time-to-live, and define an IP based rate limit. Read more here.
- Understanding the search ROI — There are no built-in analytics to tell the missing content gaps in your search database, what are your end-users unable to find, or what are the most popular searches out there and how are they performing for conversion.
Whether it’s DIY or using one of the hosted search services out there, you wouldn’t get any Search Analytics.
With appbase.io, you get analytics on popular searches, filters, results, click conversions, geography distribution, search latency. Read more here.
Our goal is to offer the best experience through the journey of developing, deploying and managing production-grade search apps. We call this the search stack! Our cumulative stack has had over 1 Million downloads till date, and tens of thousands of developers have benefitted from it in small and big ways.
Get Siddharth Kothari’s stories in your inbox
Join Medium for free to get updates from this writer.
We also strongly feel that this stack should be backed by an open core so our customers and users are never locked in. And we win customers by building the best user experience. If you think about it, the best stacks that lasted through time like SQL have only done so by being an open standard.