Table of contents
Introduction
As someone who works professionally in the Web3 space, the hackathon allowed me to combine my passion with the latest developments in the AI space. All while having fun and enjoying the process along the way!
What if training a model could be as simple as uploading a CSV file or connecting it with your existing database? With MindsDB, it is that easy.
MindsDB is an open-source, automated machine-learning platform that allows users to easily train, test, and deploy predictive models on their existing data. The best part is that MindsDB has 70+ data integrations to choose from so you can get started immediately.
Application Development
The development process was straightforward thanks to MindsDB's great documentation. Documentation is one thing we could use more of as developers in general. I was very pleased when I saw that MindsDB not only documents commands, but also core AI concepts and provides links to references.
Another great DX improvement I found very helpful is the use of Monaco Editor on MindsDB Cloud. Monaco Editor powers VS Code which we all know for its great intellisense/autocomplete. I was able to generate queries with only a few keystrokes and the setup made me feel right at home!
Connecting the data was as simple as uploading a few CSV files.
Models were trained with the following snippet:
```sql
CREATE PREDICTOR {{currency}}_predictor
FROM files
(SELECT * FROM {{currency}})
PREDICT price
```
This is the result after training models for 6 currencies.
The application was built with Next.js, TypeScript, TailwindCSS and MindsDB JS SDK (which has amazing TypeScript support, another point for DX!).
Deployed on Railway.
Business Use Case
CoinSight is a useful little app that could be used as an additional sales funnel for established Web3 businesses. Price predictions are always a trendy topic in Web3 space so the business that integrates CoinSight (or a similar project) alongside its main offering could drive more prospects to its website via CoinSight and then promote main offering alongside it.
Examples of this practice:
Conclusion
In this post, we went over the creation process behind CoinSight, the cryptocurrency price predictions app powered by MindsDB. But before we finish, I'd like to highlight something about our industry.
Software solutions usually aim to please one of two audiences:
Business people / End users
Developers
MindsDB manages to please both of them in a very competitive AI space!
It was truly a joy to develop CoinSight with MindsDB and I'll keep it in mind for future opportunities. Big thanks to MindsDB and Hashnode for organizing this hackathon and giving the community a chance to explore the technology.