Frequently asked questions
Clear, straight answers to the questions we hear most: what we do, and how we can help your business do better online.
How do we recognise phishing written by AI?
Not from the text any more. Generative AI writes without errors, can imitate your jargon and a colleague’s signature and can even manage a voice recording. So rely on process, not on impressions: confirming changes to bank details through a second channel, two-factor login and the rule that nobody sends…Jak si můžu domluvit nezávaznou konzultaci?
Napište nám přes formulář na této stránce, zavolejte na +420 555 333 158, nebo si rovnou rezervujte 15minutový hovor v kalendáři. Ozveme se do jednoho pracovního dne, na patnáctiminutovém hovoru probereme, co chcete řešit, a do pěti pracovních dnů dostanete písemný návrh řešení s cenou. Konzultace ani návrh nic nestojí…How much does training or fine-tuning an AI model cost?
The price depends on the scope of the task, the size of the model and the amount of data we work with – fine-tuning a smaller model for a specific task costs something different from training a larger model on domain-specific data. During the initial consultation we go through your…Can we run AI agents on top of a local model?
Yes, and it is one of the main reasons to go local. An agent running non-stop and processing every incoming e-mail or document consumes large volumes in the cloud and its bill is hard to predict. On your own model it runs for the price of electricity. Designing and building…How long does the audit take?
For a smaller company, as a rule two to three weeks from the start. For larger organisations the time depends on the number of departments and on the availability of the people we need to talk to.Vyřeší trénování modelu to, že si AI vymýšlí?
Samotné dotrénování ne, halucinace vznikají z podstaty jazykových modelů. Spolehlivost se řeší jinak: napojením modelu na vaše dokumenty a databáze (RAG), aby odpovídal z ověřených zdrojů, a důkladným testováním na reálných příkladech. Finetuning k tomu přidá správnou terminologii a formu odpovědí. Obojí kombinujeme podle toho, co úloha vyžaduje.How long before I see the first results?
Fixes to measurement and PPC show within weeks. SEO, content and visibility in AI answers need months, because trust and indexing have to be built. That is why we plan for 90 days and in the first month do the things with the fastest effect, so that there is something…Do links also influence who ChatGPT mentions?
Indirectly, yes. Models draw on sources that can be traced on the web and are rated as trustworthy. When you are written about on sites that carry weight, the chance grows that AI will cite you as an example of a supplier or a source of information. A mention without…Does Mediatoring provide media monitoring?
No. The company sold its media monitoring business in 2022 and now focuses on online marketing, SEO, AI optimisation and web and app development.Why do our e-mails end up in spam?
Most often because of the technical set-up of the domain. Without SPF, DKIM and DMARC records the sender looks untrustworthy. The second reason is the database: inactive addresses, contacts obtained without consent and a non-functioning unsubscribe link all damage reputation. Only in third place comes the content. An audit will…Komu se vyplatí investovat do vlastního AI modelu?
Vlastní model dává smysl firmám, které pracují se specifickými nebo citlivými daty, potřebují opakovaně řešit stejný typ úlohy – třeba kontrolu kvality ve výrobě, vytěžování dokumentů nebo predikci poptávky – nebo narážejí na limity obecných nástrojů typu ChatGPT. Pokud jde spíše o jednorázovou nebo obecnou úlohu, může být rychlejší a…What should content optimised for LLMs look like?
Content for LLMs should be factually accurate and citable, structured in a format that is easy for LLMs to digest (e.g. short paragraphs, clear headings, lists, tables), provide explicit answers to questions in natural language and contain relevant contextual metadata and structured data. Great emphasis is placed on E-E-A-T signals…
