# Open Knowledge Format (OKF) for universities | Griddo

> Griddo publishes your university’s knowledge in a format ChatGPT, Perplexity and Gemini read, understand and cite —with a link back to your page.

**URL:** https://griddo.io/en/okf-open-knowledge-format
**Language:** English

---

*Open Knowledge Format · GEO*

## If an AI talks about your university, make it talk well.

OKF publishes the knowledge your university already has on its website in a format ChatGPT, Perplexity and Gemini read, understand and cite —with a link back to your page.

---

*One file per concept*

## This is how your university’s knowledge speaks.

Every program, service or relevant fact becomes a structured file with a semantic type, a controlled description and an attribution link.

---

*People no longer just search. They ask.*

## AI already answers for your university. But it does not always answer well.

- **It hallucinates:** AI invents facts about your university: wrong enrolment figures, programs that do not exist, incorrect dates.
- **It cites others:** Third-party sources —rankings, media, forums— rank above the university itself when AI answers.
- **It ignores you:** Unstructured knowledge is left out of generative answers. AI omits it because it cannot read it.

---

*What is the Open Knowledge Format?*

## tags · attributes

OKF is a knowledge bundle —an automatic projection of the content already on your Griddo site— designed so large language models process it correctly.

Defines the concept with semantic precision. EducationalProgram, ResearchGroup, CampusService... the AI knows exactly what kind of entity it is reading.

Direct control of the message. The university decides the exact name and the description the model should associate with that concept, with no reinterpretation.

Canonical attribution link. Every file points to the official URL. When the AI cites you, it does so with a link back to your page.

Additional structured data: field, delivery mode, duration, price, language. Rich context that improves the accuracy of generative answers.

---

*Our own bundle, live*

## This is what a university’s knowledge looks like in OKF.

This is Griddo’s real OKF bundle, generated from our own schemas and published at /okf/. Every concept type —programs, events, news, taxonomies…— is declared with its frontmatter and its relationships.

---

*Generative Engine Optimization*

## GEO is the new SEO.

Where SEO optimized for showing up on Google, GEO optimizes for being cited correctly by AI. OKF is the technical layer that makes it possible.

- **01 · Accurate:** OKF publishes real, verified university data. Fewer hallucinations, more precise answers that cite your sources.
- **02 · Favourable:** Control of the narrative: the university decides what information appears, how it is described and what message it carries.
- **03 · Attributed:** AI links back to the official page. Direct traffic, brand attribution and digital authority all go up.

---

*Automatic pipeline*

## 4 · AI cites it

OKF requires nothing different from the editorial team. The process happens in the background, from content that already exists.

The team publishes in Griddo as always: programs, news, services, research. No special format, no additional effort.

Griddo reads the structured content of every page and applies the OKF engine: it selects fields, assigns semantic types and generates the files.

The OKF files are published automatically at the /okf/ path of the university’s domain, accessible and crawlable.

ChatGPT, Perplexity, Gemini and other LLMs crawl the bundle and start using that verified data in their answers, with attribution.

---

## The answer you want the AI to give.

With OKF live, generative models have correct, verified and linked data. The result is a precise answer that cites the university as its source.

What data master’s degrees does UTEF University offer?

UTEF offers the Master’s in Data Science (60 ECTS), an official program aimed at training professionals in data analysis, machine learning and advanced visualization. Available on campus and online, starting in October. For more information and enrolment, see the official site: university.edu/master-data-science

Source: Open Knowledge Format · university.edu/okf/

---

## We do not claim it: we measure it.

We compare how the models behave on the same question, with and without OKF live. The data speaks.

Without OKF

With OKF

Metric

Result

Data accuracy

Partial

High

Does it cite the university?

Does it link to the official site?

No

Yes

---

*Want to be an early adopter?*

## Engine ready and proven. Measurement, under way.

OKF deployment is in production and working. We are validating the impact on the models and building the GEO measurement tooling.

Ready

Under way

- **OKF engine:** The OKF file generator is operational and produces structured bundles from existing Griddo content.
- **Publishing:** Bundles are exposed at the public /okf/ path of the domain, accessible and crawlable by LLMs.
- **Validated end to end:** Full pipeline tested: from content in Griddo through to the model’s answer with correct citation.
- **GEO measurement:** AI presence dashboard under construction: citation frequency, data accuracy, attribution.

---

*Frequently asked questions*

## Everything you need to know about OKF

See the specification on GitHub

/en/journal/open-knowledge-format-machine-experience/

Read the full analysis

/en/journal/geo-future-digital-positioning/

GEO: the future of digital positioning

### What is the Open Knowledge Format (OKF)?

OKF is a format —published by Google Cloud in June 2026— that structures knowledge as markdown files with YAML metadata, so ChatGPT, Perplexity or Gemini can read it, verify it and cite it. It is not a database or a platform: it is a convention.

### Who is behind OKF? Is it proprietary?

It was published by Google Cloud (spec v0.1, Apache 2.0 licence) as a vendor-independent standard. Anyone can produce and consume OKF bundles, with no SDK and no account. Its value is in ubiquity, not ownership.

### What is an OKF “bundle”?

A bundle is a folder of markdown files where each file represents a concept —a program, a service, a critical date— with a YAML header (type, title, description, resource, tags, timestamp) and links between concepts. The result is a knowledge graph with no database behind it.

### How does OKF relate to GEO?

GEO determines whether your university is well represented in AI answers. OKF makes it possible: it publishes your knowledge in a form models can verify and cite, with an attribution link back to your site.

### Does it require extra work from my team?

Your team publishes in Griddo as always. Griddo reads the structured content, applies the OKF engine and exposes the bundles automatically at the /okf/ path of the domain, crawlable by LLMs. No special format and no tickets to IT.

### Is it available in Griddo already?

Generating and publishing OKF bundles is operational and validated end to end. We are building the AI presence dashboard. Available to universities already using Griddo.

### What is the difference between SEO, AEO and GEO?

SEO ranks your pages in Google’s organic results. AEO (Answer Engine Optimization) makes you the extracted answer — featured snippets, People Also Ask, voice assistants. GEO (Generative Engine Optimization) gets models like ChatGPT, Perplexity or Gemini to cite you as a source in their answers. Griddo solves all three from the same editor: native technical SEO, automatic schema markup for AEO, and OKF plus llms.txt for GEO.

### Does Griddo’s schema markup improve my visibility in ChatGPT or Perplexity?

Schema.org markup is used by Google to generate featured snippets and AI Overviews. External generative models —ChatGPT, Claude, Perplexity— ignore it, as various industry experiments show. That is why Griddo generates, alongside schema, content in OKF and llms.txt: formats designed specifically so those models read and cite your information.

### What are Google’s AI Overviews, and do I need a separate strategy to appear in them?

AI Overviews are answers generated by Gemini inside Google Search, combining several cited sources. They are not an independent channel: they appear mainly on pages already in the organic top 10 with correct schema markup. The technical SEO and automatic schema Griddo handles out of the box are the foundation for competing for that space.

---

## Make your university’s knowledge readable —and citable— by AI.

OKF is available to universities already using Griddo. Request a demo and we will show you what your institution’s bundle looks like.

Want to understand the standard in depth? Read

/en/journal/open-knowledge-format-machine-experience/

Open Knowledge Format: the standard AI reads

---

## Want to define your university’s OKF bundles?

Get in touch and we will work through how to structure your university’s ontology so LLMs describe it as precisely and accurately as possible.

---

*Generated from [Open Knowledge Format (OKF) for universities | Griddo](https://griddo.io/en/okf-open-knowledge-format)*