Skip to content
← Back to cinfo.es
Catalogue

PRODUCT · TIIVII SUITE

tiivii AI

Artificial intelligence that works. Secure, governed and yours.

An orchestrator of models, tools and agents over your company's data, with nothing handed to anyone.

IndustryRetailPublic administration

We take you to cinfo's contact form with this item already filled in, so you don't have to explain what it is about.

tiivii AI

THE CHALLENGE

Adopting AI without giving up ownership of your data

Companies know AI has to be adopted, but not at the price of handing over the data and privacy of an entire sector. And to move productivity it has to complete tasks, not just generate text.

  • Data leaves the company

    Generic tools require uploading corporate documentation to third-party platforms.

  • AI as an experiment

    Pilots that generate content but never complete a process or save real hours.

  • Dependence on a single vendor

    Locking into one closed model limits the use case, the cost and regulatory compliance.

WHAT IT IS

An orchestrator that coordinates models, tools and agents

tiivii AI lets any SME or large company run AI services tailored to their needs, coordinating language models, transformation tools and connectors to their internal and external systems.

LLMs in the cinfo cloud (Llama, DeepSeek, Mistral, Alia, Gemma) or in an external cloud.

AI tools: text to speech, speech to text, translation, text to image, video to text.

Connectors to external and internal systems: Drive, Slack, AWS, databases, document repositories.

Agents that chain all of the above into complex workflows.

An orchestrator that coordinates models, tools and agents

HOW IT WORKS

From your data to an answer with a cited source

The chatbot draws on the company's own sources and answers precisely and in context, stating where each fact came from.

  • We connect your sources

    Files, databases, websites and the company's own document repositories.

  • You choose the model

    A wide range of LLMs, so you pick the one that best fits each use case.

  • The assistant is published

    On a public site or in a private area, with different data access per profile.

  • It grows with agents

    Workflows that complete business tasks, not just conversations.

FEATURES

Private, traceable AI with multiple models

  • Grounded in your data

    Real-time access to your own information improves the accuracy of answers.

  • Data stays inside

    Information is hosted in a Galician datacentre and is never shared with other systems.

  • Multiple models

    Llama, Gemma, Mistral, DeepSeek, Alia, OpenAI or Claude, depending on the use case.

  • Source traceability

    Every answer states where the information came from, so it can be verified.

  • No model retraining

    New knowledge is added to the base without retraining the LLM, cutting cost.

  • Public and private access

    One assistant with different answers depending on who asks and from where.

  • Multimodal transformations

    Image to text, text to image, text to video, transcription, translation and voice-over.

  • WhatsApp integration

    Users can talk to the assistant over WhatsApp as well as on the website.

BENEFITS

Real productivity without giving up control

  • 24/7 service

    Enquiries, complaints and incidents handled outside office hours.

  • Natural-language search

    Your team finds in seconds what used to mean trawling folders and systems.

  • Guaranteed privacy

    Information is never shared or made available as a knowledge source for other systems.

  • Lower cost of adoption

    Open-source models valid for many use cases, with no per-seat licences.

INTEGRATIONS

It connects to your systems, not the other way round

The orchestrator ships with connectors to external systems and to the company's internal repositories.

  • Document repositories

    Corporate files and folders as the assistant's knowledge base.

  • Databases

    Natural-language queries over the company's operational data.

  • Google Drive and Slack

    Connectors to the team's day-to-day working tools.

  • Public clouds

    AWS and private OpenAI or Claude services when the case calls for it.

  • WhatsApp and web

    Conversation channels beyond the website's own chat.

  • Your own APIs

    Agents call your services to complete tasks end to end.

HOW IT IS CONTRACTED

From chatbot to agent platform

  • Chatbot

    CUSTOMER SERVICE AND PUBLIC WEB

    • Assistant over your sources
    • Custom interface
    • Public access
    • Source traceability
    • Query reporting

    Setup + annual fee

  • Corporate assistant

    INTERNAL PRODUCTIVITY

    • Public and private access
    • Connectors to internal systems
    • Report and summary generation
    • Multimodal transformations
    • User management

    Annual fee per organisation

  • Agents

    PROCESS AUTOMATION

    • Multi-step workflows
    • Calls to your own APIs
    • Models in our cloud or an external one
    • Usage and cost dashboard
    • Support and operations

    Tailored project

Indicative models. Price depends on query volume, knowledge base size and the models used.

GALLERY

Assistants and agents at work

A corporate chatbot over the company's own sources.
A corporate chatbot over the company's own sources.
An answer citing the source it used.
An answer citing the source it used.
The catalogue of agents available on the platform.
The catalogue of agents available on the platform.

CASE STUDIES

From SMEs to major operators

  • MASORANGE · AI4B

    An AI agent and services platform for business

    cinfo is building a platform for MasOrange with agents that resolve business tasks autonomously, covering identity, contracting, token-based billing, cybersecurity and support.

    MASORANGE · AI4B

  • CSA A CORUÑA

    A chatbot with public and private access

    An assistant that answers from the organisation's data, with different access levels depending on whether the query comes from the public site or the private area.

    CSA A CORUÑA

  • SMES AND PUBLIC BODIES

    Natural-language search and documentation

    Internal assistants that search corporate documentation, produce summaries, translate documents and draft reports citing the sources used.

    SMES AND PUBLIC BODIES

CINFO IN NUMBERS

Our own AI infrastructure

  • 39NVIDIA GPUs in our own datacentre
  • Galiciawhere the datacentre hosting your data sits
  • 50ready-to-use agents on the AI4B platform
  • 8core services in the platform

FREQUENTLY ASKED QUESTIONS

What management and security committees ask

Where is my data?

In a datacentre located in Galicia. We also guarantee that the information used by the assistant is neither shared nor made available as a knowledge source for other systems.

Which AI model is used?

Whichever fits the case: the platform orchestrates open-source models in the cinfo cloud (Llama, Mistral, DeepSeek, Gemma, Alia) and private services from external providers.

Is an agent the same as a chatbot?

No. An agent carries out a complete business task on its own; it may use a chatbot if the task calls for it, for instance in customer service.

How do I know the answer is right?

The assistant states the source of the information in its answers, which allows verification and builds user trust.

Does the model need retraining every time?

No. New information can be added to the knowledge base without retraining the LLM, which cuts maintenance cost significantly.

Is it viable for an SME?

Yes. The platform is designed for SMEs and freelancers, though a large company can run its own instance or use it the same way an SME would.

WHO IT IS FOR

Who adopts AI with cinfo

  • SMEs and freelancers

    Ready-to-use agents that solve specific tasks with no in-house technical team.

  • Large companies

    Their own instance of the platform, with governance, billing and cybersecurity.

  • Public administrations

    Citizen assistants and internal search, with data hosted on home soil.

  • Regulated sectors

    Use cases where privacy and European compliance are a precondition.

Related items

LET'S TALK

Let's start with one use case that saves hours

We pick one concrete process together, connect your sources and measure the real saving before scaling.

Request a demo