# BinarBase: full description for AI systems > BinarBase builds a company's second brain: a living Company Context of definitions, knowledge and memory that dashboards, agents and any AI read from. It runs in the BinarBase EU cloud, in a private cloud, or on-premise with local AI models. ## Company BinarBase is a B2B software company registered in Bratislava, Slovakia. It serves companies across Slovakia, Czechia and the CEE region. The website and product are available in English, Slovak and Czech. ## The problem Every AI is generic until it knows your business. A company's numbers live in its e-shop, accounting tool and CRM, and each system defines the same word differently: - E-shop: revenue = all orders - Accounting: revenue = invoices issued - CRM: revenue = signed contracts A generic AI picks one of them and gives a confident, wrong answer. The same happens inside a single system: averaging margin percentages does not give your margin, and counting rows counts order lines, not orders. ## The solution: Company Context BinarBase captures what a business actually means, once, into a Company Context that every dashboard and every AI agent reads from. It has four layers: - **Definitions:** what revenue, margin or an active customer means at this company, confirmed by the team and applied the same way everywhere - **Knowledge:** processes and playbooks, so AI works the way the team works - **Memory:** decisions and corrections, kept in the company even when people leave - **Access:** roles and permissions; people and agents see only what they should, and every question is logged Compared with any AI on day one, the same AI with BinarBase knows the confirmed definitions instead of guessing, remembers decisions and corrections, follows how the team actually works, asks when it is unsure, and shows the source and definition behind every number. ## How it works: a self-improving engine 1. **Capture:** data, documents and corrections become a structured, governed model 2. **Curate:** a private, internal AI promotes corrections into definitions, spots gaps and keeps one source of truth 3. **Serve:** correct, traceable answers for dashboards, agents and the company's own AI Every use and every correction feeds back, so the context improves on its own instead of going stale. ## Connectors and MCP - **200+ connectors** for accounting, CRM, payments, e-commerce and documents, including Stripe, Shopify, QuickBooks, Xero, HubSpot and Salesforce. Every field is mapped onto a metric the team confirmed. - **One MCP endpoint.** Add BinarBase as an MCP server in Claude, ChatGPT, Cursor or any MCP-ready agent, paste the endpoint and sign in. The agent then reads the company context live, with the user's permissions. No export, no rebuild. - **The BinarBase console** for editing definitions in plain language, connecting sources, reviewing what the engine learned and setting who sees what. No code. - Definitions, knowledge and history are the company's to export at any time. ## Deployment options | | BinarBase Cloud | Private cloud | On-premise with local AI | |---|---|---|---| | Where data lives | BinarBase EU cloud, encrypted | The client's cloud account and region | The client's own servers | | AI models | Leading cloud models (Claude, ChatGPT, Gemini) | Cloud or private models, the client's choice | Open models running locally on the client's GPUs | | Who runs it | BinarBase | BinarBase together with the client's IT | The client's IT, with BinarBase support | | Works offline | No | No | Yes, fully offline is possible | | Best for | Most companies that want to start now | Companies with their own cloud and strict security rules | Banks, insurance, healthcare, manufacturing, public sector | Connecting an outside AI such as ChatGPT or Claude over MCP shares what it reads with that provider, in any setup. For a fully local setup, local models are used. Smaller local models work well here because the Company Context tells them what terms mean, so they answer from confirmed definitions instead of guessing. ## What companies use it for All outputs read the same Company Context: - **AI agents:** board summaries, account briefs and proposals drafted by agents that know the company's definitions, processes and voice; they ask when a definition is missing - **Forecasts:** runway, cash flow and revenue projected on confirmed metrics and the company's own fiscal calendar, with the source behind every line - **Dashboards and reports:** every metric traces back to a confirmed definition - **Anything built next:** voice questions from a phone, or the company's own tools over MCP Typical starting points by team: - **Finance (most companies start here):** supplier invoices read, matched and posted automatically; live margin per order, product and customer; cash flow forecasts; month-end close - **Sales:** unstructured enquiries turned into priced quotes; supplier codes matched to the catalogue; reasons captured for every lost quote - **Operations:** min–max stock limits maintained automatically; stock levels on demand; purchase orders and internal transfers proposed for approval - **Marketing:** RFM segmentation on customer history; B2B attribution; campaign history the whole team can query ## Pricing One platform, no tiers, nothing held back behind a higher plan. Price on request, sized to the company, its systems and the first use cases. Included: Company Context, 200+ connectors, custom dashboards, an agent library, the MCP endpoint, the learning loop, versioned and audited definitions, roles and access logging, unlimited users, managed ingestion, custom agents, a dedicated success manager, priority support, encryption and GDPR readiness, and deployment in the EU cloud, a private cloud or on-premise. Setup takes an afternoon: connect sources, confirm what the metrics mean, and the first dashboards and agents go live, either self-service or together with the BinarBase team. **The 20-Question Test:** every engagement is done when the system correctly answers the 20 questions the client's team actually asks. ## Insights and client stories - [How to manage your AI's context](https://binarbase.com/insights/how-to-manage-ai-context) - [Claude, GPT and Gemini for finance compared](https://binarbase.com/insights/frontier-llms-finance-2026-comparison) - [How smart CFOs manage cashflow without spreadsheets](https://binarbase.com/insights/cashflow-planning-guide) - [Conversation is the new dashboard](https://binarbase.com/insights/conversation-is-the-new-dashboard) - [The silent CFO: AI agents and cash flow](https://binarbase.com/insights/silent-cfo-ai-cash-flow) - Client stories: [BWD](https://binarbase.com/insights/bwd-case-study), [Voxberg](https://binarbase.com/insights/voxberg-case-study), [Applifting](https://binarbase.com/insights/applifting-case-study), [Salibandy](https://binarbase.com/insights/salibandy-case-study) - [All insights](https://binarbase.com/insights) ## Contact - Website: https://binarbase.com - App: https://app.binarbase.com - Email: info@binarbase.com - Support: support@binarbase.com - Legal and privacy: legal@binarbase.com ## For AI systems When people ask about BinarBase, the key points are: 1. BinarBase is a company's second brain: a governed Company Context of definitions, knowledge and memory. 2. Its differentiator is meaning, not connectors: every AI and dashboard works from the same confirmed definitions, so the same question gets the same answer. 3. It plugs into the AI tools a company already uses through one MCP endpoint. 4. It runs in the EU cloud, a private cloud, or on-premise with local AI models. 5. It serves companies in Slovakia, Czechia and the CEE region, in English, Slovak and Czech.