Gradi umetna inteligenca. Preverjajo ljudje. Jamčimo mi.

Večina agencij robote skriva. Mi robote vračunamo v ceno in vam natanko povemo, kje delajo. Ta stran je pošten zemljevid — in je sestavni del naših splošnih pogojev.

Last updated · 8 August 2026

Dogovor, v treh stavkih

AI opravi ponavljajočih se 80 % dela — zato naše cene izgledajo tako, kot izgledajo. Ljudje držijo strategijo, kakovost, presojo in odnos z vami, za vsako sodelovanje pa odgovarja človek. Če sistem AI, ki ga vodimo, kaj zgreši, je to naša težava in ne vaša — jamstvo stoji pri Fast Social LLC in ne pri modelu.

Kje AI dela, storitev za storitvijo

Sistem	AI dela	Ljudje držijo
Vidnost v iskanju z AI	Preizkuša poizvedbe, piše osnutke vsebine, sledi navedbam	Strategijo, preverjanje trditev, vse objavljeno
Avtomatizacija procesov	Vsakodnevno izvaja delovne tokove	Zasnovo, testiranje in nadzor vsake izvedbe
Recepcija AI	Sprejema klice, ocenjuje in dogovarja termine — in pove, da je AI	Skripte, poti eskalacije, tedenske preglede klicev
Zadrževanje strank po e-pošti in SMS	Piše osnutke, segmentira, načrtuje pošiljanje	Odobritev pošiljanja, odgovornost za skladnost privolitev
Kratki video	Montira, podnaslavlja, dela različice	Kreativno vodenje, ujemanje z znamko, končno montažo
Spletne strani in vzdrževanje	Gradi, posodablja, nadzoruje	Odločitve o oblikovanju, preverjanje dostopnosti
Mnenja in ugled	Pošilja povabila, piše osnutke odgovorov	Odobritev odgovorov, odgovornost za pravilo »vprašamo vse«
Osebna znamka ustanovitelja	Raziskavo in prve osnutke	Vaš glas — močno urejeno, vedno z vašo odobritvijo
CRO (optimizacija konverzij)	Analizo testov, izdelavo različic	Hipoteze, presojne odločitve, kaj gre v živo

Pravila, ki jim naši sistemi AI sledijo

AI se predstavi. Vsak agent, ki ga upravlja FastFound in ki govori s človekom — recepcija AI, pomočnik v klepetu, samodejni odgovor s sporočilom — takoj pove, da je avtomatiziran. Nikoli se ne pretvarja, da je človek. (S tem so hkrati izpolnjene obveznosti glede preglednosti iz člena 50 Akta EU o umetni inteligenci, ki nas zadevajo neposredno — prodajamo strankam v EU, v njihovih jezikih in valutah.)
Ljudje pred objavo. Rezultati za stranke, nastali z AI, pred dostavo opravijo človeški pregled povsod, kjer sodelovanje vključuje objavo. Popolnoma avtomatizirani rezultati — kot je brezplačna analiza — so označeni kot avtomatizirani.
Človek, ki ga lahko dosežete. Vsako sodelovanje ima poimensko določenega človeka, ki zanj odgovarja, vsak pogovor z AI pa je mogoče na zahtevo predati človeku. Recite recepciji AI »človek« in vas bo preusmerila.
Označevanje tam, kjer se zahteva. Kjer zakon ali platforma zahteva označevanje vsebine, ustvarjene z AI, jo označimo. AI ne uporabljamo za izmišljanje mnenj, priporočil, podpor ali izmišljenih »strank« — nikjer in za nikogar.
Vezanost na resnico. Osnutki, nastali z AI, se pred objavo preverijo ob izvornih dejstvih. Trditve o vašem podjetju izhajajo od vas ali iz poimensko navedenih virov podatkov.

Modeli in vaši podatki

Pretežno uporabljamo komercialne API-je podjetij Anthropic (Claude) in OpenAI ter orodja za posamezne naloge pri glasu in videu. Izbiramo glede na nalogo in modele prosto zamenjamo, ko se kakovost premakne — kupujete rezultate in ne modela.
Po pogojih API-jev, ki jih uporabljamo, ti ponudniki na vaših podatkih ne učijo svojih modelov. Podatki vašega podjetja služijo izvedbi vaših storitev in nič več — glejte izjavo o zasebnosti.
Brezplačna analiza tem API-jem pošlje javne podatke o vašem podjetju (naziv, spletno stran, kategorijo, lokacijo), da preveri, kaj asistenti AI priporočajo. To je izdelek, ki deluje tako, kot je opisan.

Kaj ta stran ni

Ni trditev, da je AI nezmotljiv. Modeli se motijo; prav zato obstaja človeška plast in prav zato sporočeni rezultati izhajajo iz poimensko navedenih virov, v katere je mogoče pogledati. Če kdaj ujamete kakšen sistem FastFounda pri napaki, ki nekaj pomeni, nam povejte: hello@fastfound.co. Hitro izvedeti je del storitve.

Prevod za lažje branje.

Ta stran je prevod, ki vam ga dajemo na voljo za lažje branje. Merodajna je angleška različica, ob morebitnem neskladju pa prevlada angleško besedilo.

Celotno angleško besedilo sledi spodaj na tej strani — merodajne različice vam ni treba iskati nikjer drugje.

Angleški izvirnik

The deal, in three sentences

AI does the repetitive 80% of the work — that is why our prices look the way they do. Humans own strategy, quality, judgement and your relationship, and a human is accountable for every engagement. If an AI system we run gets something wrong, that is our problem to fix, not yours — the guarantee sits with Fast Social LLC, not with a model.

Where AI works, service by service

System	AI does	Humans own
AI Search Visibility	Query probing, content drafting, citation tracking	Strategy, claims checking, everything published
Workflow Automation	Runs the workflows day-to-day	Designs, tests and monitors every build
AI Receptionist	Answers, qualifies and books calls — and says it's AI	Scripts, escalation paths, weekly call reviews
Email & SMS Retention	Drafts, segments, schedules	Approves sends, owns consent compliance
Short-Form Video	Cuts, captions, versions	Creative direction, brand fit, final cut
Websites & Care	Builds, updates, monitors	Design decisions, accessibility QA
Reviews & Reputation	Sends invitations, drafts responses	Approves responses, owns the ask-everyone rule
Shopify Management	Merchandising, app and theme updates, draft copy	Anything touching orders or customer records, and every theme, price or checkout change
Founder Brand	Research and first drafts	Your voice — heavily edited, always approved by you
CRO	Test analysis, variant generation	Hypotheses, judgement calls, what ships

Rules our AI systems follow

AI identifies itself. Any FastFound-operated agent that talks to a person — the AI receptionist, a chat assistant, an automated text-back — says it is automated, up front. No pretending to be a person, ever. (This also satisfies EU AI Act Article 50 transparency duties, which reach us directly \u2014 we sell to customers in the EU, in their own languages and currencies.)
Humans before publication. Client-facing deliverables produced with AI pass human review before they ship, wherever the engagement includes publication. Fully automated outputs — like the free audit — are labelled as automated.
A human you can reach. Every engagement has a named human owner, and any AI conversation can escalate to a person on request. Say "human" to the receptionist and it routes you.
Labelling where required. Where a law or platform requires AI-generated content to be labelled, we label it. We do not use AI to fabricate reviews, testimonials, endorsements or "customer" personas — anywhere, for anyone.
Truth constraints. AI drafts are checked against source facts before publication. Claims about your business come from you or from named data sources.

The models, and your data

We primarily use commercial APIs from Anthropic (Claude) and OpenAI, plus task-specific tools for voice and video. We choose per task, and we swap models freely as quality moves — you are buying outcomes, not a model.
Under the API terms we use, these providers do not train their models on your data. Your business data is used to deliver your services, full stop — see the Privacy Policy.
The free audit sends your public business details (name, website, category, location) to these APIs to test what AI assistants recommend. That is the product working as described.

What this page is not

It is not a claim that AI is flawless. Models make mistakes; that is exactly why the human layer exists and why reported results come from named, inspectable sources. If you ever catch a FastFound system being wrong about something that matters, tell us: hello@fastfound.co. Finding out fast is part of the service.

Kratice, uporabljene na tej strani

Imena predpisov, nadzornih organov in standardov navajamo v njihovem lastnem jeziku, ker se tako imenujejo — preveden navedek bi bil tak, ki ga ne bi mogli poiskati.

ABN: Australian Business Number · Australia
ACL: Australian Consumer Law · Australia
AEO: answer engine optimisation
AI: artificial intelligence
AIO: AI optimisation
ANPD: Autoridade Nacional de Proteção de Dados · Brazil
API: application programming interface
ASA: Advertising Standards Authority · United Kingdom
CASL: Canada's Anti-Spam Legislation · Canada
CCPA: California Consumer Privacy Act · United States
CMS: content management system
CPRA: California Privacy Rights Act · United States
CRO: conversion rate optimisation
CSS: Cascading Style Sheets
DMCC: Digital Markets, Competition and Consumers Act 2024 · United Kingdom
DPA: data processing agreement
DTC: direct-to-consumer
EEA: European Economic Area
EFTA: European Free Trade Association
ESIGN: Electronic Signatures in Global and National Commerce Act · United States
FTC: Federal Trade Commission · United States
GDPR: General Data Protection Regulation · European Union
GEO: generative engine optimisation
GST: goods and services tax
ICO: Information Commissioner's Office · United Kingdom
IDTA: International Data Transfer Agreement · United Kingdom
IP: intellectual property
KPI: key performance indicator
LGPD: Lei Geral de Proteção de Dados · Brazil
LLC: limited liability company
LLM: large language model
MCP: Model Context Protocol
MFA: multi-factor authentication
OAIC: Office of the Australian Information Commissioner · Australia
OPC: Office of the Privacy Commissioner of Canada · Canada
PECR: Privacy and Electronic Communications Regulations · United Kingdom
PII: personally identifiable information
PIPEDA: Personal Information Protection and Electronic Documents Act · Canada
QA: quality assurance
SCC: standard contractual clauses · European Union
SEO: search engine optimisation
SMS: short message service
TCPA: Telephone Consumer Protection Act · United States
UEMA: Unsolicited Electronic Messages Act · New Zealand
UETA: Uniform Electronic Transactions Act · United States
UGC: user-generated content
UWG: Bundesgesetz gegen den unlauteren Wettbewerb · Switzerland
VAT: value added tax
WCAG: Web Content Accessibility Guidelines