Developers build databases to hold information. This data may be as simple as a set of username, email address pairs to complex, location based COVID-19 symptoms, contraction date, treatment length as so forth.
Assembling the data, however, is only the beginning. The data has to be queried effectively to make it useful. For instance, on a dataset of the pandemic cases, the analyst might want to know how many people under a certain age in a particular area had contracted the virus & were hospitalized. These types of queries are called SQL queries.
The thing is that database queries have to be structured correctly, the syntax has to be perfect. It’s not always easy for developers to get the results they’re after. Most SQL languages are very similar, but they’re not identical.
Kiko, a natural language translation tool, uses OpenAI’s GPT-3/Codex (Generative Pre-trained Transformer 3 is an autoregressive language model that uses deep learning to produce human-like text.) to assist developers in their query efforts.
Kiko turns questions asked in plain English, into complex queries. The tool allows developers to query their databases effectively so that their syntax is always perfect & they can get the results they’re after. This saves a lot of headaches for database developers.
The tool works for both, SQL & Panda 2, of the most commonly used databases. Additionally, the tool actually analyzes the data & using AI, suggests a selection of queries against the information that might be useful, this is incredibly beneficial & will be most useful to database programmers.
The tool is free for developers with 5 or fewer datasets to query, & allows them to translate 25 queries per month. Thereafter, you need to move on to a paid plan on Kiko. To get started, developers need to upload their dataset & start asking questions, that’s all, a very useful tool indeed.
Image credit: Kiko
