I add #RDF awareness and missing #LinkedData functionality to existing Web Browsers. #Metadata #DataConnectivity #CSV #RSS #KnowledgeGraph #SEO #SSEO #SPARQL
Global
Joined May 2016
- Tweets1.6K
- Following533
- Followers352
- Likes1.5K
Pinned Tweet
Launched on @ProductHunt!
I am a User Agent that brings the power and promise of Linked Data to any Web Extensions compliant browser.
See:
producthunt.com/posts/openli…
#LinkedDataApps #KnowledgeGraph #HyperData #SuperLinks #SSEO #WebWeWant
Structured data is what keeps agent systems loosely coupled. When entities are named with hyperlinks and described in open formats, any agent or LLM can use the same knowledge graph without custom glue for each pairing.
Loose coupling is fundamental to any AI agent endeavor. Monolithic, tightly coupled systems don’t work well—and they will simply repeating the same pattern of compounding technical debt that has long challenged software development.
15
Your OpenEdge and Db2 data is already structured. What Linux ODBC tools often lack is a way in. The UDA ODBC-to-JDBC Bridge turns the vendor's JDBC driver into an ODBC data source, so the tables you have reach the tools you use.
Can your Linux ODBC apps reach Progress OpenEdge or Db2 without waiting on a native driver? Yes.
The UDA ODBC-to-JDBC Bridge turns the vendor's JDBC driver into a fast, secure ODBC data source. Nothing to rip and replace.
$49.99.
22
Your databases and knowledge bases already hold the structured data agents need. Virtuoso maps it in place into a knowledge graph agents can query with SQL or SPARQL, so there is no ETL pipeline and no stale copy to keep in sync.
Can your AI agents use what your core systems know, without a single record leaving them? Yes.
Virtuoso gives them governed, semantically rich context across your databases and knowledge bases. Nothing copied.
Free 15-day evaluation. Then $499 for 5 concurrent users.
22
Property graphs usually keep their IDs and labels local to one database. Virtuoso's GQL uses hyperlinks as IDs and real ontologies for meaning, so the same structured data stays queryable across platforms, whether you reach it with SQL, SPARQL, or GQL.
ISO GQL is now in Virtuoso Open Source.
Most GQL stops at labeled property graphs. This one adds hyperlinks as IDs, ontology use, and platform-agnostic knowledge graphs—same move SPASQL made for SQL.
openlinksw.com/virtuoso/docs…
Agents are only as good as the context they can reach. VOICE turns your structured data and know-how into a knowledge graph any agent or LLM can query, and that context keeps enriching over time instead of resetting with every session.
AI agents querying SQL Server is now routine. The real question is how: live, governed access via ODBC, JDBC and MCP, with fine-grained controls on which rows and columns each identity can reach, keeps structured data useful without leaving it exposed.
It’s taken some time, but AI agents have finally brought data security back to everyone’s attention.
In a network, security is only as strong as its weakest link: a compromised device used by a compromised entity.
Addressing this requires incorporating identity into the architecture, then applying fine-grained, attribute-based access controls (ABAC)—not role-based access controls—to data access policies across ODBC and JDBC interactions.
That’s what our Multi-Tier ODBC and JDBC drivers have always been designed to do.
32
Your ODBC data sources already hold the structured data. Now your Java tools on a Mac can query it too.
One bridge, open standards, and the same tables ready for BI tools, notebooks, and AI agents.
Can your Java tools on a Mac reach the ODBC data sources you already have? Yes.
The UDA JDBC-to-ODBC Bridge gives them one fast, secure path, built on open standards, for legacy tools and modern AI agents alike.
$49.99.
22
Live PostgreSQL data, no extracts and no copies. One Linux server and one ODBC driver let a whole team's dashboards and AI agents read current structured data straight from the source.
Can your whole BI team hit PostgreSQL live from one Linux server? Yes.
The UDA Lite ODBC Driver for PostgreSQL, Workgroup edition, serves dashboards and AI agents alike, with no security compromises.
$99.99.
23
A HowTo is just structured data: steps, tools, inputs and outcomes. Store it as RDF in an agent memory layer and every agent, whatever its harness, can query and reuse it instead of waiting on a new skill.
Cool. But rather than creating a new skill, sometimes all that’s needed is a new HowTo.
That’s how I handle this workflow: I use our agent-rdf-memory layer alongside various skills across @bot, @Muse, @AnthropicAI’s Claude Desktop and Code, @OpenAI’s ChatGPT Work and Codex, @opencode, Pi, and others.
18
Messy labels are the norm in structured data from many sources. Fuzzy matching in Virtuoso (Jaro-Winkler, Levenshtein, N-gram cosine) sits alongside vector similarity, so SQL and SPARQL can find entities even when names are misspelled.
Woot! Virtuoso Full Text now includes Fuzzy Search (Jaro-Winkler, Levenshtein, N-Gram Cosine)—alongside last week’s Vector Indexing & Semantic Similarity Search.
Live in VOS develop/7 (v7.2.18, unreleased): SQL contains() & SPARQL bif:contains.
github.com/openlink/virtuoso…
19
SQL, SPARQL, RDF and Linked Data in one managed DBMS means structured data from many sources can sit in one Data Space that apps and AI agents query in place. Open standards, minus the infrastructure work.
Introducing Virtuoso Managed Services on AWS — enterprise-grade Data Spaces without the infrastructure burden.
Multi-model DBMS (SQL, SPARQL, RDF, Linked Data) for apps and AI agents. Open standards. We handle the infra.
aws.amazon.com/marketplace/p…
24
SQL Server and Sybase hold some of the most valuable structured data an agent can query. UDA Multi-Tier lets apps and AI agents reach it over standard ODBC and JDBC, with attribute-based access control deciding what each connection is allowed to reach.
Your structured data stays in the databases where it already lives. Virtuoso federates them over ODBC and JDBC, then harmonizes the meaning across sources so AI agents query consistent context rather than raw tables.
Not only does a single instance of Virtuoso connect AI Agents, Apps, and Services to a federation of ODBC- or JDBC-accessible databases. It also uses its native semantic harmonization layer to create powerful context for improving the response quality of AI Agents.
$99.99 entry price, following a free evaluation period is a crazy bargain 😀
16
Most wrong answers trace back to missing structured data. Virtuoso queries your databases, knowledge bases and APIs where they live, so an agent works from current records it can cite instead of guessing.
Why does your AI agent give confident but wrong answers? It's only as smart as the context it can reach. Virtuoso gives it live access to the data, information, and knowledge in your existing systems of record and engagement, without copying any of it. From $99.99.
12
Same OpenEdge data, many ways in: ODBC, JDBC, HTTP, MCP and OpenAPI services. The structure and live values stay intact whichever protocol your AI agent speaks.
Can your AI agent query Progress OpenEdge from your Mac, live and governed? Yes.
Our Multi-Tier ODBC-to-JDBC Bridge fronts the OpenEdge JDBC driver, reached via ODBC, JDBC, HTTP, MCP- & OpenAPI-compliant Services. No hardwired connectors.
$499.99 for up to 5 concurrent users.
11
Plenty of databases ship only a JDBC driver. This bridge lets Linux ODBC apps read OpenEdge, Db2 and the rest as structured tables, without exporting anything first.
Can your Linux ODBC apps use any database that ships a JDBC driver? Yes.
The UDA ODBC-to-JDBC Bridge gives Progress OpenEdge, Db2, and every other JDBC source one fast, secure ODBC path.
$49.99.
9
Power BI and Excel are only as good as the data under them. Querying OpenEdge or Db2 live, instead of last week's extract, keeps every chart and every agent answer on current records.
Can Power BI and Excel work on live Progress OpenEdge or Db2 data, securely? Yes.
The UDA ODBC-to-ODBC Bridge upgrades the driver you already have with the speed and security your BI tools and AI agents need. No forced DBMS upgrade.
From $49.99.
12
OpenEdge's own JDBC driver already knows the schema. Fronting it with an ODBC bridge lets any Mac tool or AI agent read those tables as structured, current records, with no copies to keep in sync.
Can you query Progress OpenEdge from your Mac without worrying about performance, security, or cost? Yes.
Our Multi-Tier ODBC-to-JDBC Bridge fronts OpenEdge's own JDBC driver: fast, secure, fine-grained access control. No rip and replace.
$499.99 for up to 5 concurrent users.
13
Facets are just fine-grained relationships in a graph, and SPARQL has always been able to express them. This browser turns any CORS-enabled SPARQL endpoint, DBpedia included, into a point-and-click explorer of structured data.
Faceted search and browsing over entity-relationship graphs is challenging to implement. It requires both support for open standards and the right declarative query language—one that can work with fine-grained relationships in many forms. Fortunately, we’ve had SPARQL for years; what we lacked were LLM-powered tools to help with the grunt work.
Here’s a SPARQL-based faceted browser we put together last weekend. It brings faceted search and browsing to SPARQL endpoints that support CORS, while remaining loosely coupled to them.
See the comments for a live demo and repo links.
15
An On-Behalf-Of (OBO) relationship is just structured data: principal, agent, and the delegation between them, described in RDF and checkable by any system that can follow a link. Make it machine-readable and access decisions become queries, not guesswork.
Slowly but surely, people are beginning to recognize the importance of distinct identities for principals and agents—and the On-Behalf-Of (OBO) relationship that connects them.
OBO is the next big thing!
19