{ "brand": { "word": "ABIS", "homeAriaLabel": "ABIS - home" }, "heroVis": { "ariaLabel": "Knowledge graph preview", "chromeTitle": "knowledge_graph · revenue_eur", "chromeStatus": "● LIVE", "edgeLabels": [ "SUM", "JOIN", "ENFORCES", "READ", "WRITES", "FEEDS" ], "nodeLabels": [ "core_orders_v3", "fx_rates_daily", "pii_orders", "revenue_eur", "CFO dashboard", "Audit log", "Forecast Q3" ], "legend": [ { "label": "Source", "color": "hsl(132,68%,66%)" }, { "label": "Policy", "color": "hsl(28,68%,66%)" }, { "label": "Metric", "color": "#FFAA2D" }, { "label": "Workspace", "color": "hsl(280,68%,72%)" }, { "label": "Audit", "color": "hsl(208,68%,66%)" }, { "label": "Forecast", "color": "hsl(76,68%,66%)" } ] }, "nav": { "links": [ "Product", "Knowledge graph", "Governance", "Pricing" ], "linkHrefs": [ "#product", "#graph", "#governance", "#pricing" ], "actions": [ "Book meeting", "Request access" ], "actionHrefs": [ "https://calendly.com/hello-abis-cloud/30min", "/pages/request-access.html" ] }, "hero": { "eyebrow": "EU AI Act · Data Lineage Compliance · MCP Server Audit Trails", "headline": { "line1": "Knowledge,", "accent": "governed.", "accentCycle": [ "governed.", "protected." ], "line2": "Built for Europe." }, "lead": "ABIS is the data-lineage and AI-activity audit layer for Business Intelligence MCP Server for Claude Desktop. The gatekeeper between data and AI models - a contextual intelligence layer for company knowledge that keeps every metric, dimension, and join versioned, attributed, and traceable for the EU AI Act.", "ctas": [ "Get the Whitepaper", "Watch the 1-min tour" ], "ctaHrefs": [ "/pages/request-access.html", "#tour" ], "meta": [ { "label": "SOC 2 II", "value": "in progress" }, { "label": "EU AI Act", "value": "Art. 10 · 12 · 17 by default" }, { "label": "Self-hosted", "value": "runs on your hardware" } ] }, "trustStrip": { "label": "Trusted by", "logos": [ { "text": "Siemens Energy", "mono": false }, { "text": "Takeda Austria", "mono": false }, { "text": "DBI Analytics Germany", "mono": false } ] }, "pillars": { "eyebrow": "What ABIS is", "title": "Gatekeeper, context layer, and audit trail for AI Agents running on your hardware.", "lead": "Three functions. One desktop application. Local-first. Other tools monitor AI systems and catalog the data estate. ABIS governs what AI can actually do with that estate in real time — every data source queried, every workspace accessed, every LLM call traced back to its BI context.", "cards": [ { "number": "01 · GATEKEEPER", "title": "A gate in front of every agent to protect data.", "body": "Every query passes through ABIS first. The gate reads your semantic model, applies the policy bound to the caller, redacts what shouldn't leave the workspace, and refuses what the EU AI Act says it must.", "tags": [ "RBAC · ABAC", "PII redaction at query time", "Refusal hooks" ] }, { "number": "02 · CONTEXT LAYER", "title": "A knowledge graph that survives reorgs.", "body": "Tables, columns, measures, joins, and policies as first-class nodes. Versioned, attributed, and pluggable via MCP into any agent, dashboard, or BI tool you already own.", "tags": [ "MCP server", "Power BI · SQL · SSRS ingest", "Semantic diff" ] }, { "number": "03 · AUDIT", "title": "Compliance that ships with the query.", "body": "Every read, every model change, every policy override — attributed to a person, hash-chained, replayable. Auditor mode collapses six months of activity into a single Annex IV-ready document.", "tags": [ "Hash-chained, per-run log", "Air-gap export", "Auditor export" ] } ] }, "stack": { "eyebrow": "Stack fit", "title": "What you already have. What ABIS adds.", "lead": "Most enterprises already own — or are about to buy — an AI governance platform. ABIS does not compete with it. ABIS lives one layer down, on the BI substrate, and produces the evidence those platforms expect to exist but cannot generate on their own.", "columns": [ { "tag": "GRC / AI GOVERNANCE PLATFORM", "heading": "The control plane.", "subheading": "What this layer already does well.", "items": [ "AI system registry and inventory", "Policy library and control catalog", "FRIA and risk-assessment workflows", "Vendor and model risk scoring", "Board- and regulator-facing reporting" ] }, { "tag": "ABIS", "heading": "The data and activity layer.", "subheading": "What the control plane expects to exist — but cannot generate on its own.", "items": [ "Row-level lineage across your BI estate (Art. 10)", "Hash-chained log of every agent run, model call, and prompt (Art. 12)", "Runnable SOPs that produce QMS evidence as a by-product (Art. 17)", "Replayable graph and log state at any point in time" ], "outro": "Local-first. Signed. No data leaves the building." } ] }, "flow": { "eyebrow": "The flow", "title": "Company data in. Governed answers out.", "lead": "ABIS reads from the systems your business already runs on, builds a single versioned knowledge graph, and exposes it to any agent over MCP - no rewrites, no proprietary SDK, no data leaving the building.", "arrows": [ "Import & diff", "Expose via MCP", "Query · explain · refuse" ], "stages": [ { "leftNumber": "01 · INGEST", "leftTitle": "Connectors, not copies.", "leftBody": "Read-only links to your warehouses, dashboards, and operational stores. Schemas, lineage, and metric definitions are imported - the rows stay where they are.", "nodeIcon": "database", "nodeTitle": "Sources", "nodeSub": "Read-only connectors", "nodeTag": "14 active", "nodeList": [ { "name": "postgres · core_orders_v3", "meta": "SCHEMA" }, { "name": "tableau · fx_rates_daily", "meta": "METRIC" }, { "name": "salesforce · accounts_v2", "meta": "SCHEMA" }, { "name": "policy · pii_orders", "meta": "POLICY", "type": "policy" } ], "rightBody": "14 connectors today - Postgres, Snowflake, BigQuery, dbt, Tableau, Looker, Salesforce. Bring your own via the SDK." }, { "leftNumber": "02 · GRAPH", "leftTitle": "A semantic model with provenance.", "leftBody": "Metrics, dimensions, joins, and policies as first-class nodes. Every change is signed, attributed, and diffable line-by-line - like git, for meaning.", "nodeClass": "featured", "nodeIcon": "git-branch", "nodeTitle": "Knowledge graph", "nodeSub": "Versioned · attributed · signed", "nodeTag": "v 0.4.1", "nodeList": [ { "name": "metric · revenue_eur", "meta": "2 refs", "type": "metric" }, { "name": "metric · refunded_eur", "meta": "1 ref", "type": "metric" }, { "name": "policy · pii_orders", "meta": "ENFORCED", "type": "policy" }, { "name": "audit · who-read-what", "meta": "LIVE", "type": "audit" } ], "rightBody": "Stewards review proposed changes. Auditors replay any state from any point in time." }, { "leftNumber": "03 · MCP SERVER", "leftTitle": "One protocol. Any agent.", "leftBody": "Built-in MCP server speaks the open Model Context Protocol. Any client that supports it - Claude, Cursor, your in-house agent - connects with one line of config.", "nodeClass": "green", "nodeIcon": "plug", "nodeTitle": "MCP server", "nodeSub": "abis://workspace/cfo", "nodeTag": "stdio · sse", "nodeList": [ { "name": "tool · graph.query", "meta": "READ", "dotColor": "#90DD3C" }, { "name": "tool · metric.resolve", "meta": "READ", "dotColor": "#90DD3C" }, { "name": "tool · policy.check", "meta": "GATE", "dotColor": "#90DD3C" }, { "name": "resource · audit.log", "meta": "WRITE", "dotColor": "#90DD3C" } ], "rightBody": "Every tool call is gated by policy, signed by the requester, and written to the tamper-evident log." }, { "leftNumber": "04 · CONSUMERS", "leftTitle": "Bring your own agent.", "leftBody": "ABIS is the context layer, not the brain. Use whatever model your procurement team has approved. Or use none - and run ABIS as a governed BI tool on its own.", "nodeClass": "purple", "nodeIcon": "bot", "nodeTitle": "Consumers", "nodeSub": "MCP clients", "nodeTag": "5 connected", "nodeList": [ { "name": "Claude Desktop · Lieke V.", "meta": "STEWARD", "type": "consumer" }, { "name": "Cursor · data-team", "meta": "ANALYST", "type": "consumer" }, { "name": "CFO dashboard · readonly", "meta": "VIEW", "type": "consumer" }, { "name": "in-house agent · forecast_q3", "meta": "SERVICE", "type": "consumer" } ], "rightBody": "If a consumer asks for something the policy refuses, ABIS returns a structured explanation - never a hallucinated answer." } ] }, "whyGraph": { "eyebrow": "Why a context graph", "title": "Why AI agents are dramatically better with a context graph.", "lead": "An LLM with raw SQL access is a stranger in your data estate. A graph turns that stranger into a colleague who knows the business, the dependencies, and the rules - before it touches a single row.", "cards": [ { "number": "01 · PREFLIGHT", "title": "Simulate before the change is made. Plan smart save tokens.", "body": "Before an agent through the MCP server touches a Power BI semantic model, a dbt model, or any node upstream of fifty workbooks, ABIS walks the graph and shows exactly what would break. No more try-it-and-see refreshes. Each iteration avoided is a Tableau refresh you don't pay for and a few thousand tokens you don't burn.", "tags": [ "Downstream impact", "Dry-run changes", "Token-efficient" ] }, { "number": "02 · DRIVERS", "title": "Holistic view. Real causation. No hallucination.", "body": "The graph encodes how entities actually interact - revenue depends on price, mix, churn, and marketing spend. Ask 'how does revenue move if we shift +10% into paid acquisition?' and the answer is a causal chain across the estate, not a guess from one chart.", "tags": [ "Driver analysis", "What-if questions", "Cross-domain" ] }, { "number": "03 · GROUNDED", "title": "Structure first. Then data analysis. Rules enforced.", "body": "Agents read the graph before they read the rows. They know which tables exist, which metrics are signed, where PII lives, where to write, and which writes the policy refuses. Compliance rules layer on top - the same graph that grounds the answer enforces the rule.", "tags": [ "No hallucination", "Policy-aware", "Knows where to write" ] } ], "footnote": "This is why agents love ABIS: the graph gives them a map of the territory before they take a single step." }, "product": { "eyebrow": "The desktop app", "title": "Local-first. Your metadata never leaves the building without permission.", "address": "abis://workspace/cfo/metrics/revenue_eur", "version": "v 0.4.1 · NL01", "tourTitle": "ABIS 4-minute product tour" }, "principle": { "prefix": "The opposite of an oracle. ABIS will tell you", "emphasis": "what it cannot answer", "suffix": ", and why.", "attribution": "Founding principle · v 0.1" }, "compliance": { "eyebrow": "Compliance, by construction", "title": "Ready for the EU AI Act. Not retrofitted to it.", "lead": "ABIS was started in 2024, the year the Act was adopted. The product was designed around Article 9 risk-management, Article 10 data governance, and Article 13 transparency - not bolted on after the fact.", "actions": [ "Read the whitepaper", "On data sovereignty" ], "actionHrefs": [ "/pages/request-access.html", "#" ], "facts": [ { "key": "Region", "value": "Amsterdam · NL01", "meta": "Built in the Netherlands. Hosted nowhere unless you ask." }, { "key": "Deployment", "value": "Desktop · self-hosted · air-gap", "meta": "Single binary. No phone-home. SBOM signed and published." }, { "key": "Standards", "value": "EU AI Act · GDPR · ISO 27001 (in progress)", "meta": "High-risk AI system controls available out of the box." }, { "key": "Open", "value": "MCP-native · open semantic model", "meta": "No proprietary lock-in. Export your graph as YAML, any time." } ] }, "pricing": { "eyebrow": "Pricing", "title": "Per seat. Per month. No usage meter.", "lead": "Three tiers, all self-hosted. Pay for the seats you actually have - never for the queries they run.", "tiers": [ { "name": "Startup", "description": "For one team kicking the tyres on AI governance.", "priceNumber": "EUR 79", "pricePeriod": "/ seat · month", "features": [ "Up to 15 seats", "10 workspace", "Knowledge graph + MCP server", "Default rule pack (read-only)", "Audit log — 30 days retained", "Community support (Discord + GitHub)" ], "cta": "Start 14-day trial · No credit card", "ctaHref": "/pages/request-access.html" }, { "name": "Enterprise", "description": "For organisations that have to prove it to a regulator.", "priceNumber": "EUR 599", "pricePeriod": "/ seat · month", "featuresIntro": "Everything in Startup, plus:", "features": [ "From 25 seats, no upper cap", "Unlimited workspaces", "Custom compliance rule packs", "EU AI Act Annex IV evidence export", "Hash-chained audit log — indefinite retention", "SSO, SCIM, RBAC + ABAC" ], "cta": "Get the Whitepaper", "ctaHref": "/pages/request-access.html" }, { "name": "Sovereign", "description": "For ministries, banks, and anyone running fully air-gapped.", "priceNumber": "Custom", "pricePeriod": "", "featuresIntro": "Everything in Enterprise, plus:", "features": [ "Air-gapped install + offline update bundles", "On-prem semantic model registry", "Signed SBOM, source escrow with a notary", "Named on-call engineer — NL business hours", "Custom MCP enforcement profiles", "Annual third-party security review included" ], "cta": "Talk to us", "ctaHref": "https://calendly.com/hello-abis-cloud/30min" } ] }, "faq": { "eyebrow": "Questions", "title": "Common ground.", "lead": "If yours is not here, write to hello@abis-cloud.nl", "items": [ { "question": "Is ABIS another LLM wrapper?", "answer": "No. ABIS is a context layer and a knowledge graph that any agent can connect to over MCP. Bring your own model - or run none at all and use ABIS as a governed BI tool." }, { "question": "How do I connect it with Claude Desktop?", "answer": "Since ABIS runs locally, setup is the same as any other MCP server. After you receive the desktop app, a getting started page walks you through connecting Claude Desktop with clear step-by-step instructions." }, { "question": "Why does the ABIS context graph matter?", "answer": "Because an agent without a graph is guessing. The graph gives every agent a versioned map of your data estate: which tables exist, how metrics are defined, which joins are valid, which columns are PII, which writes the policy refuses, and what downstream artefacts depend on what. That single map enables preflight simulation of changes, real driver analysis across domains, and grounded answers that don't hallucinate - all on the same substrate the EU AI Act asks you to evidence." }, { "question": "What is preflight simulation - and why does it save iterations?", "answer": "Before an agent through the MCP server modifies a Power BI semantic model, a dbt model, or any node upstream of your dashboards, ABIS walks the graph of downstream dependencies and reports exactly what would break or change. No more 'apply, refresh, see what crashes, undo'. Each loop you avoid is a Tableau refresh you don't pay for and a few thousand tokens you don't burn." }, { "question": "Can ABIS answer 'how do I grow revenue by 8%'?", "answer": "Yes - because the graph encodes how revenue actually depends on price, mix, churn, marketing spend, and the entities behind each. Ask the agent a driver question and it traverses those dependencies rather than guessing from a single chart. You get a causal chain across the estate ('shift +10% into paid -> expected lift in new accounts -> expected revenue impact at current mix'), not a single dashboard cell." }, { "question": "Why do AI agents hallucinate less with a context graph?", "answer": "Because the agent reads structure first and data second. The graph tells it which tables exist, which metrics are signed and reviewed, where PII lives, where it is allowed to write, and which writes the policy refuses outright. Hallucination collapses when the model picks from real nodes instead of inventing table names. And because compliance rules live on the same graph, the answer that grounds the question is the same one that enforces the rule." }, { "question": "Does my data leave the desktop?", "answer": "Only if you tell it to. The default deployment runs the agent, graph, and MCP server locally. No telemetry, no phone-home. SBOMs are published for every release." }, { "question": "Can I migrate from Tableau, Looker, or dbt?", "answer": "Yes. The migrate tool reads your existing semantic model and produces a draft graph for a steward to review. Lineage and joins are preserved; metric definitions are diffed line-by-line." }, { "question": "How does ABIS handle high-risk AI systems under the Act?", "answer": "Every read and model change is signed and attributed. Auditor-mode exports an Annex IV-shaped report covering risk management, data governance, and transparency." }, { "question": "Why made in NL?", "answer": "Because data sovereignty is a stance, not a marketing line. Amsterdam is our base for Tech and creativty and environment to build something great. The netherlands is envolving to an AI Powerhouse and we will be part of it." } ] }, "cta": { "title": "Govern the meaning behind your data.", "lead": "Request access to the desktop preview. We will send you a signed binary and a one-page setup guide.", "actions": [ "Get the Whitepaper", "hello@abis-cloud.nl" ], "actionHrefs": [ "/pages/request-access.html", "mailto:hello@abis-cloud.nl" ] }, "footer": { "description": "Agentic BI Services. A European startup based in Amsterdam, building a desktop product for governed BI, knowledge graphs, and EU AI Act compliance.", "columns": [ { "title": "Product", "links": [ "Overview", "Knowledge graph", "Governance", "Pricing" ], "hrefs": [ "#product", "#graph", "#governance", "#pricing" ] } ], "bottomLeft": "© 2026 ABIS B.V. · Amsterdam", "bottomRight": "Made in the Netherlands" } }