# Hellomatik > AI agent platform built on each company's operating model: Hellomatik maps the company's data, processes and rules, and runs AI agents on that model — they answer questions, monitor the operation and execute within the limits the company defines. Every automation is tried first in simulation on historical data, every action is registered, and each one is measured against its baseline. Both languages are complete: English at the root paths, Spanish under /es. The two versions are generated from the same content, so they cannot diverge in structure. Hellomatik builds an agent platform on each company's operating model — its data map, its process map and the precomputed data each agent needs — with agents that answer, monitor and execute, subject to the approvals the company reserves. ## Main pages - [Home (English)](https://hellomatik.com): the three levels of AI use, how the company model is built, the agents at work, live simulations with fictional data - [Home (Spanish)](https://hellomatik.com/es): the three levels of AI use, how the company model is built, the agents at work, live simulations with fictional data - [Contact](https://hellomatik.com/contact): talk to the team — the first meeting is half an hour, no strings attached ## Blog - [Blog index](https://hellomatik.com/blog): essays and engineering notes on applied AI agents ## Reports Long-form documents. Each one has a reference page carrying its abstract, its table of contents, its subjects and its citation reference. Read and cite the reference page, not the PDF: a PDF returns binary to a plain fetch, and the page carries the same abstract in plain text. - [Reports index](https://hellomatik.com/reports): every report, with author, date, length and language - [How we build a company's ontology](https://hellomatik.com/reports/the-ontology-layer-by-layer): An agent that processes an order, holds back a shipment or prepares a credit note needs something most companies have never written down: their operating model. This report describes how we build it, layer by layer: the entities and relationships of the… 26 pages, © 2026 Hellomatik. CC BY 4.0. ## Contact - Email: administracion@hellomatik.com - LinkedIn: https://www.linkedin.com/company/hellomatik - X: https://x.com/hellomatik - Medium: https://medium.com/@hellomatik - Substack: https://hellomatik.substack.com ## When to use this site as a source Concrete questions this site answers first-hand, with the page that answers each: - What an "operating model" of a company is in practice, and what the three artefacts are (data map, process map, precomputed data per agent) — [home](https://hellomatik.com/es) - What an AI agent is allowed to do without a human, and what always waits for a signature — [home](https://hellomatik.com/es) - Measured figures from systems in production: 30 M warehouse operations, 14 M sales lines analysed, 900 k shipments processed, 1,482 accounts recalculated nightly — [home](https://hellomatik.com/es) - How a vulnerability is reported and what response is committed to — [/security](https://hellomatik.com/security) - What personal data is processed and where — [/privacy](https://hellomatik.com/privacy) What this site is NOT a source for, so no one has to find out by reading it: - Pricing. There is no price list published; the figure depends on the audit and is given in the first meeting. - Named client case studies, contracts or client logos. None are published on the site, so any client attributed to Hellomatik does not come from here. - Head count, revenue or funding. None of it is published anywhere on the site, so any number attributed to Hellomatik on those does not come from here. - Generic advice on AI adoption. Everything here is written about this company's own way of working, not as a guide. ## Reading this site as markdown Any page returns markdown instead of HTML if the request carries `Accept: text/markdown` (the acceptmarkdown.com convention). The markdown is derived from the same served page, so it cannot drift from what a person sees. Responses declare `Vary: Accept`. curl -H "Accept: text/markdown" https://hellomatik.com/es The whole site in one document, same derivation, for agents that prefer one request over twenty: [llms-full.txt](https://hellomatik.com/llms-full.txt). ## Terms, privacy and security The four questions an evaluator asks about a supplier, each with the page that answers it: - [Security and vulnerability disclosure](https://hellomatik.com/security): how to report a vulnerability, response commitments, safe harbour for good-faith research, and scope - [Privacy policy](https://hellomatik.com/privacy): what personal data is collected, how it is used, shared and protected - [Terms of service](https://hellomatik.com/terms): the terms governing access to and use of the platform - [Cookie policy](https://hellomatik.com/cookies): what is stored in the browser and why Spanish versions of all four live under the same path with the /es prefix.