# MINEO - Full reference

> MINEO is the AI Data Platform for your whole team: one governed platform where everyone uses AI. Business teams chat and build their own assistants in Threads with no code; data teams build notebooks, live apps and pipelines - with every model and no vendor lock-in, corporate data protected, cost under control and EU compliance.

This document is a consolidated, curated reference of the MINEO product and website, written for LLM and RAG consumption. For the short index see `/llms.txt`; for the page outline see `/llms-small.txt`; the Spanish version is `/llms-full.es.txt`.

MINEO replaces fragmented Shadow AI with one governed platform. Everyone uses every AI model - OpenAI, Anthropic, Google, Mistral, Meta and open source, via Azure, Bedrock, Vertex or Ollama - with no vendor lock-in. Corporate data never trains third-party models. Governance (permissions, audit logs), cost management (visibility, smart routing, budgets) and native GDPR and EU AI Act compliance are built in. No-code or with code, in production in days.

Main app: https://b.mineo.app - Documentation: https://b.mineo.app/documentation/

## Seats and roles

MINEO is priced by seat. There are three seat types, matched to how each person works:

- **Member** - no-code AI. Uses Threads and creates Assistants over approved data and tools. The seat for business users (sales, finance, legal, operations, support).
- **Builder** - developer. Builds notebooks, live apps and pipelines, and runs cloud dev environments. Each Builder includes cloud compute (~150 CPU-hours/month). The seat for data scientists and engineers.
- **Viewer** - consumer. Views live apps and dashboards and reads shared threads. The seat for stakeholders.

## Plans and pricing

All plan and price logic is computed from a single source of truth; the pricing page has a live calculator. Current seat prices (per seat, per month):

- **Free** - EUR 0. 1 Builder + 3 Viewers. The fastest way to start.
- **Team** - pay per seat, self-serve. Member EUR 21/mo (EUR 16.8/mo billed annually); Builder EUR 65/mo (EUR 52 annual); Viewer EUR 5.5/mo (EUR 4.4 annual). Volume discounts: 10% from 25 seats, 15% from 100, 20% from 250.
- **Enterprise** - custom, with a minimum monthly commitment (from ~EUR 195/mo) and a one-time setup fee (EUR 299). Adds SSO, granular permissions, audit, BYOK (bring your own model keys), multi-cloud and private deployment, and a 99.9% SLA.

AI usage is a separate, flexible monthly budget (an optional add-on from ~EUR 7/user/month) that works as a credit across any model and any assistant - so model spend is decoupled from seats.

Pricing page: https://www.mineo.app/pricing

## Products

Every product runs on one multi-model platform: any model, protected data, EU-compliant, no vendor lock-in.

### Threads

URL: https://www.mineo.app/threads

Chat with any AI model over your data, files, databases and connected tools. MINEO turns questions into answers, visualizations and apps - governed and model-agnostic. Threads is the entry point for business users (Members).

How it works, in three steps: (1) **connect your data** - CSV, databases, APIs, files; (2) **ask a question** in plain language and get instant analysis (SQL is auto-generated, a code interpreter handles the analysis); (3) **get answers and build** - charts, insights, and production-ready apps.

- Ask your data anything - contextual answers grounded in connected sources.
- See answers come alive - interactive charts, tables and visual analysis generated from the conversation.
- From conversation to application - turn any insight into a notebook, dashboard or live app, or launch a dev environment with VS Code.
- Acts through tools via MCP (e.g. create tasks in Salesforce or Linear), not just answers. OAuth-protected MCP servers (Notion, Linear, Atlassian and more) connect once per user, never with shared credentials.
- Context window gauge next to the model picker: compact the conversation on demand, or let MINEO summarize older stretches automatically instead of dropping them.
- Global command bar (Ctrl+K / Cmd+K) across threads, projects and assistants, with quick actions.
- Personal usage widget in the sidebar: daily, weekly and monthly consumption.
- Token efficiency: prompt caching on by default (with a system prompt that stays identical between turns so the cache hits), old tool results trimmed from the history, tokens visible per message and per conversation, and the model can be switched mid-conversation; compaction runs on a cheap model by default.
- No vendor lock-in - use the best model per task (OpenAI, Anthropic, Google and more); the AI budget is a flexible monthly credit usable across any model and assistant.

Typical use cases: sales analysis, customer insights, financial reporting, operations.

### Assistants

URL: https://www.mineo.app/assistants

Build a governed AI assistant per area or process - legal, finance, sales, support. Pick any model, connect your data and tools, and ship it without code. Everything an enterprise assistant needs:

- **Capabilities** - toggle what it can do: charts, web search, code, images - with the assistant's native image model or a dedicated one from the catalog.
- **Knowledge** - connect data sources and documents as its knowledge base.
- **Integrations** - plug in external storage and tools (Drive, Slack, GitHub and more).
- **Skills** - add specialized prompts from a catalog of 122 skills.
- **MCP** - connect MCP servers and auto-discover their tools; no auth, static token or OAuth 2.1, authorized per user.
- **Deep Research** - autonomously search, cross-check and synthesize many sources into a cited report.
- **Permissions** - control who can access the assistant and with which role.
- **Organization-wide** - promote assistants and data sources to the organization and reuse them in every project; if one references files that are not shared, MINEO blocks the promotion and names the file.
- **Generate from a name** - a magic wand writes the system prompt and up to four conversation starters from the assistant's name; assistants can be tweaked or created from the thread's assistant picker.

### Notebooks

URL: https://www.mineo.app/notebooks

More than Jupyter. MINEO Notebooks combine Python, interactive widgets, AI blocks on any model, and one-click Data Apps - collaborative and governed.

- Compose with blocks - mix code, markdown, assistant, widgets and snippets in one notebook.
- Run in an isolated worker - each session gets its own encrypted execution context, kernel state, logs and packages; install packages when needed.
- Version control and real-time collaboration built in.
- Turn notebooks into live Data Apps with data-app-ready layouts and interactive controls - deploy in one click.

Docs: https://b.mineo.app/documentation/docs/notebooks/intro

### Live Apps

URL: https://www.mineo.app/live-apps

Turn notebooks and prompts into interactive dashboards, internal tools and client-facing apps - deployed with permissions and governance built in. Framework-agnostic: run Streamlit, Gradio, Dash, Panel, Shiny and other Python frameworks (9+).

- Build with code or prompts; choose your framework; publish and share.
- Always-on app endpoints - no manual reruns.
- Custom domains, embeds and white-label branding.
- Connected to live data through your data sources.
- Control via REST API with a project token: start, stop, hibernate and status, no user session, every call audited. Hibernated apps wake on traffic or when an assistant calls them as an MCP server.

Use cases: executive dashboards, operational tools, client-facing portals, AI application frontends. Docs: https://b.mineo.app/documentation/docs/live_apps

### Pipelines

URL: https://www.mineo.app/pipelines

Chain notebooks and AI steps into reliable production workflows. Schedule with cron, trigger through a REST API, and govern every run from one platform.

- Notebook-based pipeline graph - reuse notebook logic as ordered stages (extract, transform, validate, load).
- Triggers: cron schedules and REST API (`POST /v1/pipelines/{id}/run`).
- Configurable runtime resources per pipeline; full execution history and traceability.

Use cases: ETL/ELT flows, feature engineering, model training jobs, warehouse syncs, automated reporting, app and dashboard refreshes. Docs: https://b.mineo.app/documentation/docs/pipelines

### Dev Environments

URL: https://www.mineo.app/dev-environments

A full Linux workspace in the browser with VS Code, terminal access and Git workflows - and the freedom to install Codex, Claude Code, Gemini CLI or whatever your team needs.

- Editor, file explorer and terminal, no local setup.
- Install packages, system tools, custom Docker images and GPU.
- Git-native workflow; connected to your data sources and project context.
- One workspace to build notebooks, live apps and pipelines.

Docs: https://b.mineo.app/documentation/docs/dev-environments

### Integrations

URL: https://www.mineo.app/integrations

Connect everything - data sources, storage, code, auth, notifications and any Python package.

- **Data sources:** CSV, MariaDB, MySQL, AWS Redshift, Oracle, SQL Server, PostgreSQL, ClickHouse, MCP.
- **Storage & collaboration:** Google Drive, SharePoint, OneDrive, Teams.
- **Code:** GitLab, GitHub, Git.
- **Authentication / SSO:** PingFederate, Active Directory / LDAP, ADFS, OpenID Connect, SAML, GitHub Auth, Google OAuth.
- **Notifications:** Email, Slack.
- **Extensibility:** Python packages, apt packages, Docker images, MCP servers (no auth, static token or OAuth 2.1 - Notion, Linear, Atlassian - authorized per user), plus custom integrations.

## Why MINEO

### Security and compliance

URL: https://www.mineo.app/why-mineo/security-compliance

Your data never trains a model. Every interaction is permissioned, logged and auditable. GDPR and the EU AI Act are covered natively, so the AI program isn't stuck waiting on compliance sign-off.

Four security principles:

- **Protected data** - prompts and documents never train external models; you choose what is shared and with which provider.
- **Permissions & access** - SSO, role-based access, granular permissions on assistants, data sources and documents; credentials to external tools (MCP over OAuth) are per user, never shared.
- **Traceability** - audit logs of every interaction by user, model, agent and use case, plus admin actions (remote session close, extra budget grants); organization admins can close any active session remotely. Ready for audits and regulatory reviews.
- **Governance** - common policies that end Shadow AI and bring every team into one controlled environment.

Deployment: EU-hosted managed cloud (multi-tenant SaaS); or deploy in your own AWS, Azure or GCP account inside your perimeter; or pin data residency to the region your regulations require. Data is encrypted in transit and at rest.

Note on certifications: ISO 27001 / SOC 2 are not yet held (work in progress); control is preserved via own-cloud/own-region deployment, data that never trains models, granular permissions and a full audit trail.

### Cost management

URL: https://www.mineo.app/why-mineo/cost-management

AI spend is the new shadow cost - scattered ChatGPT/Copilot licenses, untracked API keys, models priced ~30x apart used for the same task. MINEO turns scattered spend into a budget you can manage, see and optimize.

- **Visibility by team and use case** - know who consumes, how much, and which processes justify the spend.
- **Smart routing** - send simple tasks to more efficient models; reserve advanced models for high-value work.
- **Budgets and limits** - an organization ceiling, then limits per user, group and project by day, week and month; warning at 80%, block at 100%; one-off extra budget grants, all audited. Each user sees their own daily, weekly and monthly usage in the sidebar.
- **Consolidation** - reduce duplicate licenses and centralize adoption on one platform.

### No vendor lock-in

URL: https://www.mineo.app/why-mineo/vendor-lock-in

The AI market re-prices roughly every six months; a frontier model version lasts ~18 months before deprecation. MINEO is the orchestration layer that lets you use the best model per case, switch when convenient, and keep agents and apps running through every change - no rebuild.

- Choose per use case; switch providers without rebuilding.
- Use your own contracts; bring your own keys (Enterprise).
- Routing policies per case; future-proof by default.

## Compliance detail (GDPR and the EU AI Act)

- **GDPR** (EU Regulation 2016/679; Spain's AEPD): lawful processing, data minimization, right to be forgotten, portability, 72h breach notification. Fines up to EUR 20M or 4% of global turnover. MINEO provides a signed DPA, auto-generated Records of Processing Activities (ROPA), a pre-filled DPIA template (Art. 35), and verifiable deletion on request.
- **EU AI Act** (EU Regulation 2024/1689; Spain's AESIA): high-risk obligations apply from August 2026 (HR, credit, education, critical infrastructure). Fines up to EUR 35M or 7% of global turnover. MINEO provides automatic risk classification per use case, technical documentation (Art. 11), a risk management system (Art. 9), logging records (Art. 12) and a FRIA template (Art. 27).

## Solutions by industry

URL: https://www.mineo.app/solutions

Every vertical follows the same governed pattern: replace Shadow AI with one platform (permissions + audit), keep corporate data private (never trains third-party models; you control residency and access), give each team conversational AI, assistants, governed notebooks, internal apps, dev environments and automated pipelines - and stay audit-ready for that industry's regulation. The compliance framing per industry:

- **Education** (/solutions/education) - GDPR and the EU AI Act; institutional data, including minors' records, stays private.
- **Energy & Utilities** (/solutions/energy-utilities) - NIS2, critical infrastructure and GDPR.
- **Finance & Banking** (/solutions/finance-and-banking) - DORA, GDPR and the EU AI Act.
- **Healthcare & Life Sciences** (/solutions/healthcare-life-sciences) - GDPR, medical confidentiality and the EU AI Act.
- **Insurance** (/solutions/insurance) - Solvency II, DORA, GDPR and IDD.
- **Legal** (/solutions/legal) - attorney-client privilege, GDPR and the EU AI Act.
- **Manufacturing** (/solutions/manufacturing) - trade secrets, IP, NIS2, GDPR and the EU AI Act.
- **Marketing & Sales** (/solutions/marketing-sales) - GDPR, ePrivacy and the EU AI Act.
- **Media & Entertainment** (/solutions/media-entertainment) - copyright/IP, GDPR and the EU AI Act.
- **Retail & E-commerce** (/solutions/retail-e-commerce) - GDPR, PCI-DSS and the EU AI Act.
- **Transportation & Logistics** (/solutions/transportation-logistics) - NIS2, GDPR and the EU AI Act.

## Concepts (glossary)

URL: https://www.mineo.app/academy

- **AI Workspace** - a single platform where a whole company can use, build and govern AI across every model, instead of scattering it across disconnected tools. It centralizes access, data, permissions and cost. MINEO is an AI Workspace.
- **Multi-model AI** - using more than one AI provider (OpenAI, Anthropic, Google, open source) through a single platform, picking the best model per task and switching freely without rebuilding.
- **Vendor lock-in** - dependence on a single AI provider, where switching models means rebuilding agents, apps and workflows. Avoiding it means using any model through one platform and changing providers without rewriting anything.
- **Shadow AI** - unsanctioned use of AI tools by employees (e.g. pasting company data into public chatbots) outside IT's visibility and control. It creates data, cost and compliance risk; the fix is governed AI teams actually want to use.
- **AI governance** - the controls that keep enterprise AI safe and accountable: permissions, audit logs, data boundaries and policies over which models and data each team may use.
- **AI cost management** - making AI spend visible and controllable: usage per team and use case, routing each task to the cheapest viable model, and budgets and limits.
- **AI assistant (agent)** - a configured AI that performs a specific job using a chosen model, approved data and connected tools. Unlike a raw chatbot, it is governed, reusable and permissioned.
- **AI workflow** - an automated, multi-step process where AI runs on a schedule or trigger (e.g. a daily report or recurring data refresh) without manual work each time.
- **MCP (Model Context Protocol)** - an open standard that lets AI assistants connect to external tools and data sources (databases, ticketing systems, internal APIs) so they can act, not just answer.

## Company

- **About** (https://www.mineo.app/about) - MINEO is built so a whole company can use, build and govern AI on one platform. Principles: every model, no lock-in; govern, don't block; cost you can see. Partners include Microsoft for Startups and NVIDIA Inception.
- **Customers** (https://www.mineo.app/customers) - used by data scientists, data analysts, enterprises and education; trusted by teams including Air Liquide, Brambles, Heineken, Procter & Gamble and Suntory.
- **Contact** (https://www.mineo.app/contact) - talk to the team or request a demo.

## Key links

- Home: https://www.mineo.app
- Pricing: https://www.mineo.app/pricing
- Security & compliance: https://www.mineo.app/why-mineo/security-compliance
- Cost management: https://www.mineo.app/why-mineo/cost-management
- Vendor lock-in: https://www.mineo.app/why-mineo/vendor-lock-in
- Solutions: https://www.mineo.app/solutions
- App: https://b.mineo.app
- Documentation: https://b.mineo.app/documentation/