I wanted to share PlugAndData, a tool we built to close the gap between code and data.
By combining the concept of a database and a data pipeline into a single tool, data storage and interfaces can be managed automatically: no external database schema to be maintained anymore, no more object-relational mappers nor database queries. Just write code and run it, even scaled across multiple machines or with human-in-the-loop.
The user interface follows this typical visual programming approach where the user drags in nodes and connects them. There are data, function and query nodes.
An interesting detail is that we have modelled hashing/memoization simply as another node. By giving the hashing node both an input and an output port which you can connect or disconnect independently any time, hashing becomes flexible. For example, you may decide to override some hashed results by connecting only the input port while processing a given batch of data.
We built PlugAndData because we saw the challenge for small and mid‑sized teams of juggling with multiple tools from databases over orchestrators to containerization.
I’d love to hear the community’s thoughts on the concept, we are still in the early pilot phase.
There is a also a 4 minutes demo video on the site (we integrated an LLM, which makes for a nice demo but is not seen as the core benefit/selling point by us).
Hi HN,
I wanted to share PlugAndData, a tool we built to close the gap between code and data. By combining the concept of a database and a data pipeline into a single tool, data storage and interfaces can be managed automatically: no external database schema to be maintained anymore, no more object-relational mappers nor database queries. Just write code and run it, even scaled across multiple machines or with human-in-the-loop. The user interface follows this typical visual programming approach where the user drags in nodes and connects them. There are data, function and query nodes. An interesting detail is that we have modelled hashing/memoization simply as another node. By giving the hashing node both an input and an output port which you can connect or disconnect independently any time, hashing becomes flexible. For example, you may decide to override some hashed results by connecting only the input port while processing a given batch of data.
We built PlugAndData because we saw the challenge for small and mid‑sized teams of juggling with multiple tools from databases over orchestrators to containerization.
I’d love to hear the community’s thoughts on the concept, we are still in the early pilot phase. There is a also a 4 minutes demo video on the site (we integrated an LLM, which makes for a nice demo but is not seen as the core benefit/selling point by us).
Looking forward to feedback!