Graphene

Graphene은 AI 네이티브 데이터 플랫폼으로, 오픈소스로 공개된 데이터 분석 프레임워크입니다. 기존 SaaS 내 개별 에이전트들이 각기 제한된 도구와 컨텍스트만 다루는 문제를 해결하고자, 코딩 에이전트가 모든 데이터 작업을 수행할 수 있도록 설계되었습니다. Graphene은 대시보드-애즈-코드와 시맨틱 레이어를 결합해 ANSI SQL의 강력함과 전통적 거버넌스를 유지하며, 버전 관리가 가능한 개발 환경을 제공합니다. 이를 통해 데이터 전문가들이 컨텍스트 엔지니어링에 집중하고, 에이전트가 실질적 데이터 작업을 수행하도록 지원합니다. 향후 Slack 등 업무 환경에 자연스럽게 통합되어 에이전트가 데이터 질의와 시각화를 자동화하는 미래를 제시합니다.

https://graphenedata.com/blog/introducing-graphene/

#ai #dataanalytics #opensource #agent #semanticlayer

Introducing Graphene • Graphene

Announcing our new company, Graphene, and the open source launch of our namesake data analytics framework.

Show HN: SLayer, a semantic layer maintained by your agent

SLayer는 에이전트가 데이터베이스와 연결해 의미론적 계층(semantic layer)을 관리할 수 있도록 설계된 오픈소스 도구입니다. 기존 SQL MCP 서버나 BI용 정적 의미론적 계층과 달리, 에이전트가 상호작용을 통해 학습하고 의미론적 모델을 진화시킬 수 있도록 지원합니다. Python 클라이언트, 스키마 자동 생성, 스키마 변화 감지, 복잡한 다단계 쿼리 DSL 등 AI 에이전트 기반 데이터 분석에 최적화된 기능을 제공합니다. 현재 접근 제어, 캐싱 기능 등이 로드맵에 포함되어 있습니다.

https://github.com/MotleyAI/slayer

#semanticlayer #agent #database #opensource #dataanalysis

GitHub - MotleyAI/slayer: SLayer: a lightweight semantic layer for AI agents and humans

SLayer: a lightweight semantic layer for AI agents and humans - MotleyAI/slayer

GitHub

Ontopic Suite 2026.1 is here. What's new?

- Kubernetes
- arm64 and amd64 fully supported
- lot of UI polishing
- Azure Blob Storage
- updated Ontop engine

Visit https://ontopic.ai/en/activities/ontopic-suite-2026.1-released/ for more information

#knowledgeGraphs
#rdf
#rdb2RDF
#semanticlayer

🧩 Sciogli i nodi dei tuoi incubi di data engineering con il potere del semantic layer! Rinventando la gestione dei dati. #DataEvolution #SemanticLayer ✨📈

🔗 https://www.tomshw.it/business/come-costruire-chatbot-affidabili-sui-tuoi-dati-aziendali-2025-12-16

La risposta agli incubi di un data engineering è il semantic layer

I team affrontano la sfida di costruire chatbot affidabili: la soluzione è implementare un livello semantico che elimini l'ambiguità

Tom's Hardware
Does your company struggle with messy, inconsistent data? A semantic layer transforms complex data into a single, business-friendly view, enabling reliable self-service analytics and trusted AI. Find out why you need one now. #DataAnalytics #SemanticLayer #BusinessIntelligence #DataGovernance
https://inpathways.net/is-companys-data-frustratingly-useless-why-you-absolutely-need-an-essential-semantic-layer/
🤡 A riveting 21-minute #guide for those desperate to build a "semantic layer" with DuckDB—because who wouldn't want to spend their precious time wrestling with #YAML files? 🐤📚 The authors promise you’ll emerge enlightened, or at least mildly confused, about why this even matters. 🙃
https://motherduck.com/blog/semantic-layer-duckdb-tutorial/ #DuckDB #SemanticLayer #DataEngineering #TechHumor #HackerNews #ngated
Why Semantic Layers Matter — and How to Build One with DuckDB - MotherDuck Blog

Learn what a semantic layer is, why it matters, and how to build a simple one with DuckDB and Ibis using just YAML and Python | Reading time: 21 min read

MotherDuck

Refactoring #data projects helps with clarity, scalability, and onboarding new people into the project

We’re refactoring in the open by tackling our internal analytics dbt project

In part 1 we create a time spine as a foundation for semantic layer

https://medium.com/inthepipeline/building-in-the-open-recces-internal-dbt-repo-refactor-53f7860d6a1a

#DataEngineering #BuildingInTheOpen #OpenSource #Analytics #Data #dbt #SemanticLayer

Building in the open: Recce’s internal dbt repo refactor

Coming into an incomplete, nascent, or otherwise confusingly set-up codebase is a rite of passage for data folks. At my first real job, at the Federal Reserve Bank of San Francisco, I inherited a…

In the Pipeline
#KnowledgeGraphs work best when they're accompanied by a #domainKnowledgeModel. This creates a #semanticLayer where everyone in your enterprise can see and connect the data, content, and knowledge they need. Andreas Blumauer has been developing this powerful idea for 20 years.

https://knowledgegraphinsights.com/andreas-blumauer/
Andreas Blumauer: the elements of the enterprise semantic layer

Andreas Blumauer says knowledge graphs work best with a domain knowledge model, creating a "semantic layer" that maps business knowledge.

Knowledge Graph Insights

Question for #data people -- I'm trying to understand headless BI. In particular, what it solves that SQL doesn't. In other words, why dbt (pre-analytics layer) isn't enough.

From what I can tell, it solves the problem that SQL can't easily be parameterized. You can make views that slice and dice your transactional data, but those views would hardcode a bunch of decisions better left up to the consumer. You can also denormalize the heck out of your data to make all conceivable queries easy, but then you end up with an analytical table that's way too tall and wide.

It seems like what these semantic layer / headless BI tools do is apply the metadata to your transactional data that allow for BI tools to offer slick query builder interfaces, which ultimately are generating SQL. Furthermore, the logic for different types of analysis can be standardized and controlled, compared to people handwriting their queries.

Do I have this right?

#semanticlayer #analyticslayer #headlessbi #businessintelligence #analytics #dataanlytics #datamastodon

Enterprise sponsored #opensource:
ADP, a global leader in cloud based solutions for Human Capital Management, sponsored the development of “rules”, (https://github.com/ontop/ontop/pull/576), another great feature of @ontop4obda
.

#semanticlayer
#knowledgegraphs

Extend the VKG with INSERT SPARQL rules by bcogrel · Pull Request #576 · ontop/ontop

This PR introduces a major feature: using SPARQL INSERT rules to extend the VKG. This work has been sponsored by ADP, a big thank to them! The first idea for this feature came from this thread on t...

GitHub