Launched this week
yggy
The AI knowledge graph that builds itself
81 followers
The AI knowledge graph that builds itself
81 followers
Most "chat with your data" tools throw a pile of text chunks at a model and hope for the best. yggy does what they skip: it turns your cluttered documents and connected systems into a real knowledge graph, with no schema to design and no pipeline to run. Every fact keeps its source. Duplicates merge themselves, and you can undo any merge. Then one graph serves everyone: your team in the app, your code through the API, your agents through MCP. Hosted in the EU. Your knowledge stays yours.
Hey Product Hunt 👋
We're Dominik, Carlos and Max, the team behind yggy, and we're really glad to finally show it to you.
Company and personal knowledge lives everywhere: PDFs, mailboxes, drives, tickets. Most AI tools on top of that either retrieve a pile of text chunks, or let an agent search through MCP and API calls and guess. It's slow, it burns tokens, and you can't always tell where an answer came from. We wanted answers we could check. So we built yggy to turn that mess into a knowledge graph that builds itself.
How it works
Bring data in. Upload documents, connect your systems, or push data through the API.
yggy does the structuring. It reads everything, extracts entities and relations into your ontology, merges duplicates, and links every fact back to its source.
Use it your way. Ask in the app, build on the REST API, or plug the graph into your agents via MCP.
What we care about
Transparent: every answer points to where it came from, and you can see what yggy did and why
Ontology compliant: facts land in a structure we and you define, not in a pile of loose text
Dev first: REST API, MCP, a CLI and SDKs, with docs written for builders
Multi-tenant and secure: every organization and project is isolated, with SSO, scoped API tokens and an audit log
Fixable: merges are reversible, and unclear cases go to a review queue
Yours: EU-hosted, GDPR-first
Just want your own notes in a graph? We built yggy notes for exactly that.
We'd love your questions, your toughest challenges, or just a good chat. We're here all day and read every comment.
Dominik, Carlos and Max
Addition from the co-founder department:
We didn’t wake up dreaming of building another Knowledge Graph tool.
We wanted to build Knowledge Graphs ourselves. Then we discovered how much time disappears into data modelling, ontology design, entity matching and explaining to five systems that they are talking about the same thing.
By the time the graph finally worked, everyone involved had reconsidered their career choices.
That’s how yggy happened: we automated the part of graph building that normally makes people give up on graph building.
Try it, challenge it and break it. Dominik particularly enjoys reproducible bugs.
Oh this looks amazing! I am building an AI companion and keeping good memory of users is critical. Would that work or is your product more focused on enterprise?
@heychristoph thanks, really appreciate it! Our main focus is teams and companies, but the core works for a single person too!
yggy turns conversations and documents into a knowledge graph that remembers who someone is, what matters to them and how it all connects - all ontology compliant.
Your companion could plug in through our API, SDK, CLI or MCP server. Each user gets their own private space, and everything is hosted in the EU.
Very interesting, is it open source?
@rasul_osmanbayli Thanks! Parts of it are. Our graph engine, graphraum, is open source under Apache 2.0. It's the WebGL engine that draws yggy's interactive graphs in 2D and 3D.
Code: https://github.com/yggy-tech/graphraum
Docs and demo: https://yggy-tech.github.io/graphraum/
The yggy platform itself isn't open source right now. We are however planning to have yggy Community as a self-hosted open core under AGPL-3.0.
Anyways, we'd love to hear what you'd want to build with it!
Very cool launch. We have recently started experimenting with an LLM wiki in Obsidian via Codex and are amazed by what's possible. But there are a lot of details that are really hard to get right if you have to think about the whole infrastructure, agents, and maintenance yourself. Super curious to try yggy out!
keeping every fact linked to its source is what would make me trust an answer. when two duplicates merge wrongly, how do you spot it before undoing?
@domudev ah that makes sense, undoing from the Resolved tab is reassuring. do the alias links show up on the entity page itself?