DUKE AI
Duke Analytics SAS provides an EU-sovereign, agent-native analytics platform that converts natural-language questions into verified, signed decisions and audited actions. It serves European SMBs through enterprises across retail, pharma, banking, telecom, public sector, and embedded SaaS publishers.
- Company typePrivate
- Founded2024
- HeadquartersSaint Paul, United States
- Headcount1–10
- GTM typeB2B
- OfferingSoftware
What DUKE AI does
Duke Analytics SAS is a French private software company founded in 2024 and headquartered in Paris, with operational roots in La Réunion. It provides an agent-native analytics platform that closes the full data-to-action loop through four integrated modules: Text2Data (natural-language to displayed and verifiable SQL), a governed Semantic Layer with typed, versioned, human-validated certified metrics, an MCP (Model Context Protocol) server enabling agentic execution across connected systems, and an Action Layer that triggers email/Slack/Jira/Lemlist actions after human validation with cryptographic signatures and complete audit trails. The underlying Duke Agents Runtime — covering AtomicAgent (persistent memory), CodeActAgent (code generation), and L'Orchestrateur (multi-agent coordination) — was released under Apache 2.0 in July 2025 after 18 months of R&D, while the rest of the platform remains proprietary.
The company serves European SMBs, mid-market firms, and enterprises across retail, life sciences/pharma, financial services, telecommunications, wealth management, data consulting, and public sector, plus SaaS vertical publishers embedding analytics via white-label API. Named customers include Sanofi (pharmacovigilance and commercial ops), Orange Réunion (legacy telecom migration), Bank One Maurice (conversational risk reporting), MyAssetRocks, Datanalysis, Wiiv, Mipsit, and KDI. Commercialisation combines a tiered per-seat subscription (Basic €39, Professional €69, Enterprise €99 per user/month), usage-based Embedded/OEM contracts, and professional services for enterprise and regulated deployments. Approximately 60+ clients across 4 countries are reported, with distribution via self-serve PLG, enterprise field sales (Paris, London, Berlin), and API-first embedded channels.
The company's differentiator is contractual EU sovereignty: all inference runs on Mistral AI, Scaleway, and OVHcloud by default, with no data sent to US APIs unless the customer opts in. The firm is positioned as GDPR-aligned by design, SOC 2-ready, and compliant with the EU AI Act as of August 2026. It appears privately held and founder-owned — no external venture funding rounds are disclosed in available sources — with a small team distributed across Paris, London, Berlin, and remote locations.
DUKE AI firmographics
Firmographics- Name
- DUKE AI
- Legal name
- Duke Analytics SAS
- Website
- https://duke-ai.io
- Company type
- Private
- Founded year
- 2024
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- Duke Analytics SAS provides an EU-sovereign, agent-native analytics platform that converts natural-language questions into verified, signed decisions and audited actions. It serves European SMBs through enterprises across retail, pharma, banking, telecom, public sector, and embedded SaaS publishers.
- Ownership category
- akta.pro rank
Where DUKE AI is headquartered
LocationHeadquarters
- HQ city
- Saint Paul
- HQ country
- United States
- HQ region
- North America
Offices2 records
Markets served
DUKE AI business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Marketing or Sales, Infrastructure, Operations
Revenue model
- Subscription - Per User: Tiered per-user monthly subscription model with three main tiers (Basic €39, Professional €69, Enterprise €99 per user/month). All tiers include EU-sovereign hosting. Enterprise tier includes dedicated CSM and onboarding.
- Embedded / OEM API: Headless API deployment for white-label embedding in third-party SaaS products. Usage-based billing with OEM and co-marketing contracts. Dedicated Solutions Engineer provided.
- Professional Services: Custom onboarding, deployment, and implementation services for Enterprise and Embedded tiers, particularly for regulated industries and 50+ seat deployments.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Per seat | Monthly | Basic: First questions to your data at €39/user/month |
| Per seat | Monthly | Professional: Single source of truth, monitored and acting at €69/user/month |
| Per seat | Monthly | Enterprise: Full power with SaaS connectors, agents, audit at €99/user/month |
| Other | Pay-as-you-go | Embedded: Duke in your product via API - custom quote |
Go-to-market motion3 records
Distribution channels3 records
Marketing channels7 records
DUKE AI product offering
Product offeringCore offering
Duke Analytics sells an EU-sovereign, agent-native analytics platform that converts natural-language questions into verifiable SQL against certified enterprise data and triggers human-validated actions. The platform is delivered as tiered per-seat subscriptions (Basic, Professional, Enterprise), a custom-priced embedded API product for white-label deployment in SaaS applications, and an open-source agent orchestration runtime (Apache 2.0).
Product overview
Duke Analytics is a unified, agent-native analytics platform for European SMBs and enterprises that closes the full data-to-action loop — from natural language querying through governed decision-making to auditable execution. The platform is built around four core modules: Text2Data (natural language to SQL), the Semantic Layer (governed, versioned metrics with full lineage), the Action Layer (validated execution to email/Slack/Jira), and the MCP Server (agentic execution across systems). Three commercial plans (Basic at €39, Professional at €69, Enterprise at €99 per user/month) offer escalating capabilities, plus an Embarqué API product for white-label embedding. The open-source Duke Agents Runtime (Apache 2.0) provides the underlying agent orchestration framework. All inference runs EU-sovereign on Mistral AI, Scaleway, and OVHcloud.
Differentiator
Problem solved
Functional benefit
Products and services
- Duke Analytics Platform Agent-native analytics platform that turns natural-language questions into verifiable, signed decisions on certified enterprise data, closing the full data-to-action loop without leaving the platform. Delivered via tiered per-seat subscriptions (Basic, Professional, Enterprise) and used by SMBs through to regulated enterprises.
- Duke Embarqué (Embedded API) White-label API product enabling SaaS publishers to embed Duke's conversational analytics, semantic layer, and action capabilities directly inside their own products under their own UI, with usage-based billing, OEM contracts, and co-marketing.
- Duke Agents Runtime Open-source AI agent orchestration framework (Apache 2.0) including AtomicAgent (persistent memory), CodeActAgent (code generation), and L'Orchestrateur (multi-agent coordination) built on a European stack with Mistral AI, powering the company's commercial platform and available for community use.
Quantifiable outcome
- −98% time-to-insight (2-5 days → 30 seconds)
- +8 more outcomes
Companies that use DUKE AI
Customer profileNamed customers12 records
Segments10 records
Ideal customer profiles4 records
DUKE AI technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration12 records
AI capability10 records
Feature6 records
DUKE AI partnerships and signals
Strategic signalPartnerships
Twelve partnerships are on record, tiered core and minor.
- Mistral AIcoreEU LLM provider for sovereign inference. Duke routes to Mistral by default for EU-hosted inference. Mistral-large for reasoning, Codestral for code generation. 10x cheaper than US alternatives with open weights available.
- ScalewaycoreEU cloud infrastructure provider. One of Duke's GPU dedicated hosting locations in France for AI inference. Part of sovereign hosting stack.
- OVHcloudcoreEU cloud infrastructure provider. Part of Duke's sovereign hosting stack for data and inference in Europe.
- ElectrophazzminorMusic label/artists featured on Duke Analytics About page. 'Juste les vibes qu'on aime pour vos données.' Electrophazz album played in studio between sessions.
- SpotifyminorPodcast distribution partner. 'Sessions' podcast available on Spotify alongside Apple Podcasts, YouTube, and RSS.
- Apple PodcastsminorPodcast distribution partner. 'Sessions' podcast available on Apple Podcasts alongside Spotify, YouTube, and RSS.
- YouTubeminorPodcast/video distribution partner. 'Sessions' podcast available on YouTube alongside Spotify and Apple Podcasts.
- JiracoreAction Layer integration for validated action execution. Duke triggers Jira tickets after human approval in workflow.
- SlackcoreAction Layer integration for validated action execution. Duke sends notifications and updates via Slack after human approval.
- Salesforcecore200+ SaaS connectors including Salesforce. Enterprise tier includes API keys and MCP server access for major CRM integration.
- HubSpotcore200+ SaaS connectors including HubSpot. Enterprise tier includes API access for CRM and marketing automation integration.
- Stripecore200+ SaaS connectors including Stripe. Enterprise tier supports payment data integration for analytics.
Scale indicators10 records
DUKE AI competitors and assessment
Company assessmentEmerging players
- Hex: Hex is an AI-enhanced collaborative data notebook platform with conversational features for SQL, Python, and visualization - an emerging player whose agentic AI capabilities overlap with Duke's Text2Data and Action Layer positioning.
- Mode Analytics: Mode is a collaborative analytics platform combining SQL, Python, and R with AI-assisted features - an emerging player in the same conversational analytics space as Duke, focused on data team workflows rather than Duke's broader business-user PLG motion.
Broad incumbents
- Tableau: Tableau (Salesforce) is the dominant traditional BI incumbent that has added Tableau AI/Einstein capabilities for natural language querying - a broader incumbent that competes across Duke's market while also serving customers far beyond Duke's analytics-agent niche.
- ChatGPT Enterprise: ChatGPT Enterprise offers natural language data analysis on uploaded files and connected sources via Advanced Data Analysis - a broad incumbent/alternative explicitly compared by Duke in its content, and the leading general-purpose LLM competing for conversational analytics workloads.
- Google Looker: Google Looker is a BI platform with natural language and Gemini-powered AI features, part of the Google Cloud ecosystem - a broad incumbent targeting enterprise analytics customers that Duke also pursues, particularly with EU data residency requirements.
- Microsoft Power BI: Power BI is Microsoft's BI platform with Copilot AI features for natural language to DAX/SQL, deeply embedded in the Microsoft Fabric ecosystem - a broad incumbent that competes with Duke Analytics particularly in EU enterprise accounts using Microsoft infrastructure.
Direct peers
- Sisense: Sisense provides an embedded analytics platform with AI capabilities and white-label deployment via API - a direct peer competing with Duke's Embarqué/OEM offering for vertical SaaS publishers wanting in-product analytics.
- ThoughtSpot: ThoughtSpot offers AI-powered analytics with natural language search, code generation, and embedded analytics capabilities - a direct peer targeting the same conversational analytics and SaaS-embedding use cases as Duke.
- Akkio: Akkio is an AI analytics platform targeting SMBs and agencies with natural-language-to-insights workflows - a direct peer competing for the same conversational analytics customers Duke serves with its Basic and Professional tiers.
- Julius AI: Julius AI is a conversational AI data analyst that converts natural language into SQL, Python, and visualizations - directly comparable to Duke's Text2Data capability and explicitly named as a competitor in Duke's own comparison content.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat4 records
Key risks5 records
Key highlights7 records
Customer concentration
DUKE AI social profiles
Digital presenceDUKE AI compliance and trust
Trust signalCompliance3 records
DUKE AI financial estimates
Financial estimateRevenue estimate
Valuation estimate
DUKE AI leadership team
Management profileNumber of profiles
Profiles1 record
DUKE AI funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
DUKE AI M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about DUKE AI
What does DUKE AI do?
Duke Analytics sells an EU-sovereign, agent-native analytics platform that converts natural-language questions into verifiable SQL against certified enterprise data and triggers human-validated actions. The platform is delivered as tiered per-seat subscriptions (Basic, Professional, Enterprise), a custom-priced embedded API product for white-label deployment in SaaS applications, and an open-source agent orchestration runtime (Apache 2.0).
Is DUKE AI a public or private company?
DUKE AI is a private company. It is classified as founder individual operated bootstrapped and is currently operating.
When was DUKE AI founded?
DUKE AI was founded in 2024. It employs 1 to 10 people.
Where is DUKE AI based?
DUKE AI is headquartered in Saint Paul, United States, in the North America region.
How does DUKE AI make money?
Three revenue lines are on record. Subscription - Per User is the primary driver. The others are embedded / OEM API and professional Services.
Who are DUKE AI's main competitors?
Emerging players on record are Hex and Mode Analytics. Broad incumbents are Tableau, ChatGPT Enterprise, Google Looker and Microsoft Power BI. Direct peers are Sisense, ThoughtSpot, Akkio and Julius AI.
Does DUKE AI have an API?
Yes. Duke Analytics offers a headless API (Embarqué plan) providing the same capabilities within a custom UI, enabling white-labelling. Enterprise plan includes API keys and MCP server access for Claude, ChatGPT, and Mistral models. Developer documentation is at dev.duke-analytics.com.