What is jev4pg?
jev4pg is an open-source framework that adds a semantic layer to PostgreSQL. You ask questions in natural language, and it turns them into SQL while keeping the relational calculations exact. It is built around JEV, a system that evaluates semantic decisions with typed evidence.
You start by connecting your existing PostgreSQL tables and views. Then you ask a question in ordinary language, such as "Which suppliers have overdue, partially delivered orders?" The framework generates SQL you can inspect and correct. You can rerun queries from history, and JEV handles semantic tasks like filtering, classifying, and routing text.
It suits developers and data analysts who want to keep control of their SQL while adding AI-powered interpretation. The project is Apache 2.0 licensed, so you can inspect and extend it. A native PostgreSQL preview adds probability embeddings that store answer distributions for each record.
jev4pg features
- Natural language to SQL: Ask questions in plain English and get executable SQL you can review.
- Semantic operators: Filter, classify, and route text with typed JEV decisions via workspace, API, or SQL.
- Evidence registry: Reuse semantic observations and provenance across queries for consistent results.
- Review before writing: Preview inserts, updates, and deletes before committing changes to your data.
- Probability embeddings: Store answer distributions for fixed questions and compare records locally.
What you can do with jev4pg
- Query supplier performance with natural language questions.
- Classify customer feedback into categories with typed decisions.
- Extract structured entries from text with source evidence.
- Compare records by semantic similarity using probability embeddings.
How to get started with jev4pg
- Install jev4pg using Docker Compose or the installation guide.
- Connect your PostgreSQL tables and views to the workspace.
- Ask a question in natural language, inspect the generated SQL, and run it.
Tips for better results with jev4pg
- Connect your PostgreSQL tables and views first, then ask questions in plain English. Review the generated SQL before running it to catch misinterpretations early.
- Use jev4pg's semantic operators for filtering, classifying, and routing text. Store decisions in the evidence registry so you can reuse consistent observations across queries.
- Preview any inserts, updates, or deletes before committing changes. jev4pg lets you inspect the exact SQL, so verify the operation matches your intent.
- Rerun queries from history to refine your questions. Correct the interpretation once, then reuse it to save time and keep results consistent across analyses.
jev4pg pricing
jev4pg’s homepage does not list prices. Check jev4pg.com for current plans.
What to check before you rely on jev4pg
- Check whether your PostgreSQL version supports all jev4pg features, especially the probability embeddings preview, since the website does not list version requirements.
- Verify how jev4pg handles your data privacy. The site does not state whether queries or evidence are sent to external LLM services or stay local.
- Confirm the export formats for results and evidence. The website does not specify whether you can export beyond SQL output or integrate with your existing tools.
Who jev4pg is for
jev4pg suits data analysts, developers and data engineers. If that is not you, the AI data analysis tools below may fit better.
Similar tools compared with jev4pg
| Tool | What it is | Pricing |
|---|---|---|
| jev4pg | Open-source natural language SQL framework | See website |
| Julius AI | AI workspace for data and content | See website |
| Rows | AI-powered spreadsheet and data analyst | See website |
| Clay | AI platform for go-to-market workflows | Freemium |
| Semrush Copilot | SEO and AI search visibility platform | Freemium |
jev4pg FAQ
What is jev4pg?
jev4pg is an open-source framework that adds a semantic layer to PostgreSQL. It lets you query your data in natural language while keeping SQL for exact calculations. It is built around JEV for semantic decisions.
Is jev4pg free to use?
The website does not state pricing details. It is open source under Apache 2.0, so you can inspect and extend the code. Check the GitHub repository for the latest information.
Who is jev4pg for?
It is for developers and data analysts who want to combine natural language queries with SQL control. You can attach existing PostgreSQL tables and views without copying data.
What can I build with jev4pg?
You can build semantic layers on top of PostgreSQL, such as natural language query interfaces, text classification pipelines, and evidence-based data extraction. The framework includes a web interface, HTTP API, and Python runtime.
Does jev4pg connect to my existing PostgreSQL tables and views?
Yes, jev4pg is designed to work with your existing PostgreSQL setup. You connect your current tables and views, then ask questions in natural language. The framework generates SQL you can inspect and correct. The website confirms this workflow, so you can integrate it without migrating your data.
Can I use jev4pg to classify customer feedback or extract structured entries from text?
Yes, jev4pg includes semantic operators for filtering, classifying, and routing text. You can classify customer feedback into categories with typed decisions and extract structured entries from text with source evidence. These features are part of the framework's core capabilities, as shown in the official use cases.
Does jev4pg require an account or a paid plan to use?
The jev4pg website does not mention any account requirements or pricing. The project is open-source under Apache 2.0, and you can explore it on GitHub. For details about setup or any potential paid services, check the official documentation or the GitHub repository.
