Reducing search response time by 99.98% with Xata

Parker Dinkins, a Florida-based entrepreneur and self-taught developer, launched his startup, Precision Ecology, in 2022 with a group of friends. The company focused on digital mapping and data solutions, targeting sustainable development and environmental management. They used a Wingtra drone for gathering aerial images over large tracts of land. Parker's responsibility in processing these images led to his deep interest in geospatial data and GIS.

When Precision Ecology didn't pan out, Parker switched gears and went back to school to dig into Environmental Studies and focus on GIS.

So what does an experienced entrepreneur like Parker do when he’s back at school? Launch another startup, of course.

If you’re unfamiliar with the GIS world, it is full of disparate data sets, layers, shapes, and point files. Every industry, agency, and region tends to have its own collection of open-source assets available for download and use. This is great because the data is accessible, but navigating through various web entities to find what you’re looking for can be a bit of a pain.

Cue OpenSourceGISData.com (OSGD), Parker's latest venture. OSGD is a website that contains geoportals and ArcREST servers for all the major federal agencies, states, counties, and cities. After seeing some popularity on Reddit and healthy daily traffic, he feels he scratched an itch for geospatial professionals.

About a year after launching the OSGD website, he officially launched a sister search engine GISData. GISData has indexed over 740,000 layers from 1,700 different sources and counting. The goal was to provide a single source of truth where users can search across all disparate data sources. And so far, it’s proven to be a hit.

Balancing straight A’s at the University of Central Florida with a budding startup is no easy task, but somehow Parker is managing it just fine.

In the initial release of the search engine, Parker was pushing CSV directly to the client and used JavaScript code to search against it. This lead to many issues with timeouts and lag, and the search took a long time. He was OK with this, but knew he wanted to get the initial release out quickly given the interest from the GIS community.

For a while, he tried MongoDB, but got lost in their documentation and struggled to implement full-text search for his use case. After a bit more searching (no pun intended), he discovered Xata as an alternative to MongoDB. Paired with Vercel, he was able to build out the initial powerful geospatial data search experience in a day. Since then, it’s been very easy to continue to iterate and focus on the pipeline with an ever-growing data set.

The search behind GISData is built on a modern technology stack. It’s currently using Dagster for orchestration, Python for indexing scripts, Xata as the database, JavaScript and UIkit for the frontend and Vercel for application deployment. This stack is optimized for speed and minimal operational maintenance.

Parker has a larger vision for this platform beyond the GISData search engine. He’s planning to roll out a free and paid version of the GIS API that enables developers to integrate geospatial data discovery into their own applications. For less technical users, he’s aiming to create a simple mapping platform where users with little to no experience can find data and create simple maps for their own personal use case. We also discussed the possibility of introducing an AI geo assistant, which is another avenue of monetization. While he’ll likely be focusing on the next few months on foundational elements to support those features, there’s a bright future ahead for the Omniverse.xyz GIS platform.

After talking about what’s on the horizon, we asked Parker to share some of his favorite parts about Xata and where he thinks we can improve. Here are some of the reasons why Parker chose Xata for his application:

• Reducing time to market. The simplicity and developer experience Xata provided significantly reduced the time it took to take the GISData search implementation to production.
• Zero database management. Databases can be tricky and complex. Because Xata is serverless, there is zero-ops needed.
• More intelligent search. Parker is interested in using the Xata ask endpoint to augment his current search experience with chat like experiences.

When asked what he’d like to see from Xata, here’s what Parker shared:

• Dagster integration. Using Dagster for orchestration, he’d love to see a native integration similar to Airbyte.
• URL rich data type. While he currently has URLs stored as strings, it would be useful to have a data type that stores URLs and checks their validity.

Do you have a similar story or community contribution you'd like to share? Send us an email or ping us on Discord if you'd like to be featured in our community spotlight. Until then, happy building 🦋