# Ortelian

> Ortelian combines an AI platform with deployment services to put agents into business processes end to end, on a world model of the business and the world it operates in.

Public pages are available as HTML and as Markdown. Request `Accept: text/markdown` on a page URL, or use the `.md` URL listed next to each page. Canonical Source URLs in those files point at the HTML page.

**What Ortelian does**

Ortelian makes companies AI-native: agents run business processes end to end, within the goals and limits the team sets, so the company gets more done, faster, with the team it already has. Ortelian takes a business from experimenting with AI to running entire processes with it.

- Agents take on ongoing work autonomously. They follow what is happening in the company and its market, investigate what matters to a goal, and act in the company's systems or involve a person. People set direction and decide where agents need approval.
- Agents work from a shared, persistent world model of the business and the world around it: companies, people, facilities, markets, events and how they relate, connected to the records, conversations and observations that describe them. Findings, methods and work records from agent work are retained, so what one agent learns can inform what another does next. Agents use that picture to test explanations and decide what to do next, even when no single source spells out the answer.
- Deployment services put agents into use: an on-site audit, process redesign from first principles around what AI can now do, connections to the company's systems, agent workflows, evaluations and training. Ortelian then hosts and operates the platform, monitors the work and keeps improving it with the customer.
- Evaluations built on real examples of the work check that agents meet the standard agreed with the team. Quality, completion time and human effort are tracked, and the evaluations are rerun as the system changes.
- Access and actions are configured for the work. Tools are granted to agents explicitly, approval points are defined before a workflow goes live, and execution is recorded.
- Teams give tasks, assign ongoing work and shape how it gets done through Ortelian's AI interface, Slack and email.
- Ortelian works across AI models. Models are tested against the customer's work while the business context, knowledge and working methods stay in the platform. Customers own their data and world model and can export both; default provider arrangements do not use customer data for training.
- Over time the company's world model, evaluations and workflows compound: the team can delegate more, shape workflows themselves and take on work it could not do before.

**Who Ortelian is for**

Companies in Europe and the US with complexity in their team, product, service or market, and a clear ambition to become AI-native. Ortelian works best with teams ready to move quickly, give feedback and put improvements into practice. The first area is chosen with the customer's team, where AI can make a meaningful difference, such as sales, finance or operations.

**How an engagement runs**

- A two-week on-site audit maps the work, identifies the constraints, redesigns the process and sets the first implementation scope and a baseline for measuring results.
- Delivery then runs in two-week cycles: scope, build with the team, evaluate against real examples, deploy what meets the standard, review and set the next scope.
- The first engagement is about ninety days on one part of the business. Ortelian stays to operate the platform and improve the work.
- To start, Ortelian needs an internal owner with the authority to make decisions and drive implementation, ideally a founder or someone on the leadership team, and a team willing to test, give feedback and iterate. Implementation also requires time with the people doing the work and agreed access to the relevant systems and data.
- Pricing covers deployment services and ongoing platform use. Scope, expected usage, support and fees are agreed before work begins. No public price list is published.

To discuss fit, use the [contact form](https://ortelian.com/contact/) or email laurens@ortelian.com. For a customer example, see [One week at one company](https://ortelian.com/notes/one-week-at-one-company/).

**How an agent should use this site**

Fetch the relevant page with HTTP GET and `Accept: text/markdown`, or fetch its linked `.md` file directly. For example: `curl -H 'Accept: text/markdown' https://ortelian.com/platform/`. Use the page's canonical Source URL when citing it. To start an engagement, direct the user to the contact form or prepare an email describing the process, systems involved, internal owner and desired outcome; send it only at the user's request.

This website describes Ortelian and publishes its notes. It is not a chatbot, a public API or an MCP server, and it offers no programmatic access to the Ortelian platform. Access to the platform is part of a customer engagement. No public pricing or self-service execution endpoint is available.

Public pages may be crawled, indexed, summarized, and cited. Name Ortelian as the source and link to the canonical Source URL. Do not present Ortelian writing as your own or republish substantial portions as a substitute for linking. Respect robots.txt and reasonable request rates.

## When to use Ortelian

- [Go-to-market investigation](https://ortelian.com/index.md): Use when a team needs to track companies and market changes, investigate emerging needs and decide where to focus. [HTML](https://ortelian.com/).
- [Finance reconciliation](https://ortelian.com/index.md): Use when recurring work involves reconciling records across systems, investigating discrepancies and preparing exceptions for human review. [HTML](https://ortelian.com/).
- [Operations workflows](https://ortelian.com/platform.md): Use when work spans systems and agents need shared business context, ongoing investigation and agreed limits on actions or escalation. [HTML](https://ortelian.com/platform/).
- [Process audit and deployment](https://ortelian.com/how-we-work.md): Use when a company needs help choosing and redesigning a process, connecting its systems, testing agents against real work and putting them into daily use. [HTML](https://ortelian.com/how-we-work/).
- [Discuss a first engagement](https://ortelian.com/contact.md): Use when the user wants to assess fit or scope implementation; provide the contact form or email rather than attempting an API call. [HTML](https://ortelian.com/contact/).

## Company

- [Home](https://ortelian.com/index.md): Deploying AI inside complex companies: agents that run business processes end to end. [HTML](https://ortelian.com/).
- [About](https://ortelian.com/about.md): Company approach and founder Laurens Nys. [HTML](https://ortelian.com/about/).
- [How we work](https://ortelian.com/how-we-work.md): AI platform plus deployment services: on-site audit, redesign, integration and launch, then ongoing platform operation. [HTML](https://ortelian.com/how-we-work/).
- [Platform](https://ortelian.com/platform.md): Agent software, a world model of the business, evaluations and action limits. [HTML](https://ortelian.com/platform/).
- [Contact](https://ortelian.com/contact.md): Start a conversation about the work you want AI to take on. [HTML](https://ortelian.com/contact/).

## Notes

- [Notes](https://ortelian.com/notes.md): The latest essays on agents, company knowledge and putting AI to work. The links below follow our suggested reading order. [HTML](https://ortelian.com/notes/).
- [The model can be rented. The world it works in cannot.](https://ortelian.com/notes/the-model-can-be-rented.md): Every company rents the same intelligence at a falling price. The advantage that lasts is the maintained model of your own business that the intelligence works in. [HTML](https://ortelian.com/notes/the-model-can-be-rented/).
- [Why coding agents work and company agents don't](https://ortelian.com/notes/why-coding-agents-work-and-company-agents-dont.md): Coding agents took off first because a repository is already a world an agent can stand in. A company is not, so you have to build that world before asking the agent to work. [HTML](https://ortelian.com/notes/why-coding-agents-work-and-company-agents-dont/).
- [What is a world model for a company?](https://ortelian.com/notes/what-is-a-world-model-for-a-company.md): A world model is a live, linked model of what exists in a business and the world around it, with provenance on every fact, that agents work from instead of raw text. [HTML](https://ortelian.com/notes/what-is-a-world-model-for-a-company/).
- [One week at one company](https://ortelian.com/notes/one-week-at-one-company.md): A week after the first call, a sales team works from a list that refreshes itself, a meeting-prep agent they check before every call and a territory model a new rep can ask for a list from. [HTML](https://ortelian.com/notes/one-week-at-one-company/).
- [Redesign the work before you automate it](https://ortelian.com/notes/redesign-the-work-before-you-automate-it.md): An agent on the current process makes the current process faster, waste included. So the first two weeks of a deployment redesign the work, and only then do we build. [HTML](https://ortelian.com/notes/redesign-the-work-before-you-automate-it/).
- [Agree the standard before you build](https://ortelian.com/notes/agree-the-standard-before-you-build.md): The deployments that last agreed quality, cost, time and human effort on real examples before anything was built. The pilots that die never agreed what good meant. [HTML](https://ortelian.com/notes/agree-the-standard-before-you-build/).
- [Access is not a map](https://ortelian.com/notes/access-is-not-a-map.md): Connecting an agent to the CRM, Slack and email gives it territory. The map, what those things are and how they relate, still has to be built. [HTML](https://ortelian.com/notes/access-is-not-a-map/).
- [More context makes agents worse](https://ortelian.com/notes/more-context-makes-agents-worse.md): Past a point, adding context degrades an agent. It needs the smallest sufficient view of the work, and only a model of the company can decide what that view is. [HTML](https://ortelian.com/notes/more-context-makes-agents-worse/).
- [What is an ontology?](https://ortelian.com/notes/what-is-an-ontology.md): An ontology says what exists in a business, what each thing means, how things relate and what can be done with them. The graph then holds the particular companies and people. [HTML](https://ortelian.com/notes/what-is-an-ontology/).
- [What sight means](https://ortelian.com/notes/what-sight-means.md): Connected facts show what no single source contains. That is the real payoff of a world model, and the larger bet we are making. [HTML](https://ortelian.com/notes/what-sight-means/).
- [The case for GraphRAG](https://ortelian.com/notes/the-answer-is-between-the-documents.md): The case for GraphRAG is a broad view of the evidence that can be reused. Better coverage than basic retrieval, without processing every document for every question. [HTML](https://ortelian.com/notes/the-answer-is-between-the-documents/).

## Optional

- [Why the name](https://ortelian.com/1570.md): The name comes from Abraham Ortelius's 1570 atlas, not from an incorporation year. [HTML](https://ortelian.com/1570/).
- [Privacy](https://ortelian.com/privacy.md): How the public website handles data. [HTML](https://ortelian.com/privacy/).
- [Terms](https://ortelian.com/terms.md): Terms of use for the public website. [HTML](https://ortelian.com/terms/).

- [Full text](https://ortelian.com/llms-full.txt): core commercial pages and published notes in one Markdown file.
- [Sitemap](https://ortelian.com/sitemap.xml)
- [Notes feed](https://ortelian.com/feed.xml)

## Contact

- [Email Ortelian](mailto:laurens@ortelian.com): Discuss the process you want to improve.
- [Contact form](https://ortelian.com/contact.md): Send a few lines about your business and what you want AI to take on. [HTML](https://ortelian.com/contact/).