What ellamind does

ellamind GmbH is a Bremen-based private German software company that builds a full-lifecycle platform for evaluating, deploying, and monitoring AI agents in regulated industries. Its core products are elluminate, an agentic evaluation platform offering criteria-based scoring, experiment comparison, and quality gates before production; ellarun, an agent runtime providing credential brokering, sandboxed execution, and EU AI Act Article 12 audit trails by default; elluminate live, a real-time security layer with three-tier pattern, intent, and conversation policy enforcement; and ellaverse, a portfolio of domain-specific simulated evaluation environments with verified ground truth (live: PKV Claims Processing for German private health insurance and Retailer Product Listing QA; in development: Commercial Underwriting, KFZ Liability Claims, Legal Document Review). The company is model-agnostic, integrates with Claude, GPT, Mistral, Llama, and customer-owned models, and runs on German data centers under EU data protection standards.

The company sells to enterprise customers in German health insurance (hkk Krankenkasse, BITMARCK, an unnamed GKV anchor customer), insurance more broadly, banking and finance, public sector, and regulated e-commerce, with a freemium plus sales-assisted model that escalates from a permanent free tier through Pro subscription to multi-year Enterprise contracts featuring SSO, on-premise or air-gapped deployment, and dedicated support. Distribution combines direct enterprise field sales with channel partnerships (BITMARCK, T-Systems, ITZBund) and EU-funded research consortium participation (OpenEuroLLM, LLMs4EU, SOOFI, LLM4KMU). A separate Frontier Research line co-funded by EU and federal programs delivers bespoke vertical-agent development at the state of the art.

The company is privately held by its two co-founders, Dr. Jan Philipp Harries (CEO) and Björn Plüster (CTO), whose prior open-source German-language LLMs have been downloaded over one million times on Hugging Face. The team of approximately 14 named members includes multiple PhD-level AI researchers; the company is actively hiring across engineering, AI, product, and sales and is positioned to capitalize on the August 2026 EU AI Act GPAI enforcement and December 2027 high-risk system deadlines that create urgent demand for audit-ready AI deployment infrastructure.

Name

ellamind

Legal name

ellamind GmbH

Website

https://ellamind.com

Company type

Private

Founded year

2024

Operating status

Operating

Short description

ellamind GmbH builds a full-lifecycle platform for evaluating, deploying, and monitoring AI agents in regulated European industries. Its elluminate, ellarun, elluminate live, and ellaverse products serve German health insurers, banks, public agencies, and large e-commerce operators with EU AI Act-ready audit trails and on-premise deployment.

Ownership category

akta.pro rank

Product category

AI Agent Evaluation & Deployment Platform

NAICS

Computer Systems Design and Related Services (5415), Custom Computer Programming Services (541511), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821)

SIC

Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373), Services-Computer Programming Services (7371)

akta.pro primary industry

End-to-End MLOps & ML Platform Suites (HDAAABAA)

akta.pro secondary industries

AI Observability, Monitoring & Evaluation Platforms (Drift, Quality, Safety) (HDAEANAF), Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC), AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ), AI/ML Solution Integration & MLOps Enablement (BPAEAAAJ)

Keywords

• AI agent evaluation platform
• Enterprise AI deployment
• EU AI Act compliance software
• Regulated industries AI agents
• AI agent observability

Headquarters

HQ city

Bremen

HQ country

Germany

HQ region

Europe

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. elluminate SaaS Subscriptions: SaaS subscriptions for elluminate evaluation platform with tiered plans (Free, Pro, Enterprise). Pricing is value-based, oriented to team size, evaluation volume, and deployment requirements. Professional and Enterprise plans include unlimited evaluations, with Enterprise offering SSO, on-premise/air-gapped deployment, and dedicated support.
2. ellarun Agent Runtime (Enterprise): ellarun agent runtime deployed as part of Enterprise engagements. Pricing is work-in-progress; sales team provides tailored deployment plans. Deployment modes include cloud, on-premise, or air-gapped.
3. Frontier Research — Custom Development: Custom AI development engagements for frontier research projects, co-funded by EU and federal programs. Development at the state of the art for clients requiring bespoke vertical agents.

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels6 records

Core offering

ellamind builds a software platform that enables enterprises in regulated industries to evaluate, deploy, and monitor AI agents across their lifecycle. The stack comprises the elluminate agentic evaluation platform (offline criteria-based scoring, quality gates, EU AI Act compliance reporting), ellarun agent runtime (sandboxed execution, credential brokering, audit trails), elluminate live real-time monitoring, ellaverse domain-specific simulation environments, and a Compliance Documentation Assistant covering EU AI Act, GDPR, DORA, and DSA.

Product overview

ellamind is a German AI company offering a comprehensive platform for building, evaluating, deploying, and monitoring AI agents for regulated industries. The core product ecosystem centers on elluminate (AI evaluation platform) and ellarun (agent runtime), with built-in monitoring (elluminate live) and simulated testing environments (ellaverse). elluminate provides criteria-based scoring, experiment comparison, and quality gates before production, while ellarun handles secure deployment with credential brokering and audit trails. ellaverse offers domain-specific evaluation environments for healthcare (PKV claims), insurance, retail, and legal use cases. The Compliance Documentation Assistant integrates with elluminate to cover EU AI Act, GDPR, DORA, and other regulatory requirements. All products are designed for regulated industries including health insurance, banking, and public sector, with EU-native architecture and data sovereignty by default.

Differentiator

Problem solved

Functional benefit

Brands

• elluminate: Agentic evaluation platform for offline criteria-based scoring, experiment comparison, and quality gates before production.
• ellarun
• ellaverse
• elluminate live

Products and services

• elluminate Agentic evaluation platform for AI reliability. Provides offline criteria-based scoring, experiment comparison, quality gates before production, version-tracked prompt templates, test collections, binary/pass-fail criteria, conversation evaluation, agentic trajectory evaluation, RAG evaluation, cost and latency analytics, a Python SDK and REST API, and an Eval Agent that builds or analyzes evaluations from natural language descriptions. Includes pre-built EU AI Act compliance packages and generates audit-ready PDF reports with criterion-level breakdowns.
• ellarun The agent runtime for deploying any AI agent to production with security, credential brokering, and full audit trails. Features isolated sandbox execution, declarative permissions, secrets management so agents never see raw API keys, EU AI Act Article 12 audit trails by default, and cloud, on-premise, or air-gapped deployment modes.
• elluminate live Real-time security layer that monitors every agent action, enforces policies written in plain language, and blocks dangerous operations instantly. Three-tier architecture comprising Tier 1 deterministic pattern rules, Tier 2 LLM-based intent analysis against natural-language criteria, and Tier 3 conversation analysis with every action logged and auditable.
• ellaverse Simulated realistic workspaces for developing and evaluating AI agents without affecting production data. Live environments include PKV Claims Processing (German private health insurance GOÄ billing rules, 27 cases, 283 decisions, 18 error types) and Retailer Product Listing QA (33 cases, 33 decisions, 7 error types). Additional environments in development: Commercial Underwriting, KFZ Liability Claims, and Legal Document Review.
• Compliance Documentation Assistant Guided workflow for AI compliance documentation covering EU AI Act, GDPR, DORA, and DSA. Covers risk classification, Annex IV requirements (63 fields across 14 task types), generates audit-ready PDF reports with criterion-level breakdowns and evaluation evidence. Built-in pre-built compliance packages cover EU AI Act prohibited practices and Fraunhofer IAIS assessment dimensions.
• Frontier Research — Custom Development Custom AI development engagements for frontier research projects, co-funded by EU and federal programs. Development is performed at the state of the art for clients requiring bespoke vertical agents. Outputs include open-source artifacts published on Hugging Face and GitHub such as propella-1, sui-1, base-eval, and inference-hive.

Quantifiable outcome

• Case correctness doubled from 40% to 93% on a 30-claim GOÄ fee schedule evaluation using the same model with structured domain instructions
• +4 more outcomes

Named customers4 records

Segments6 records

Ideal customer profiles6 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration6 records

AI capability12 records

Feature5 records

Partnerships

Nine partnerships are on record, tiered flagship, core and minor.

• OpenEuroLLMflagshipStrategic or Co-development Partnerellamind is part of a 20-member consortium of leading European research institutions, companies, and EuroHPC centers developing a family of high-performance multilingual foundation models for commercial, industrial, and public applications. The consortium aims to democratize access to high-quality AI and strengthen Europe's competitiveness through transparent, compliant open-source models.
• LLMs4EUcoreStrategic or Co-development Partnerellamind is a partner in LLMs4EU, contributing to the development of state-of-the-art language models with European languages, values, and innovation at the center. This EU-funded consortium pools expertise from across Europe.
• SOOFI (Sovereign Open Source Foundation Models)coreStrategic or Co-development PartnerA German consortium of research institutions and startups developing open and sovereign AI language models as a European alternative to existing systems. SOOFI aims to develop a high-performance open-source foundation model aligned with European values and regulatory requirements from the outset.
• LLM4KMUminorStrategic or Co-development PartnerA consortium of leading research institutions, companies, and innovation partners in North Rhine-Westphalia making large language models accessible to SMEs. The project provides a shared experimentation platform, shared know-how, and prototypical use cases to help SMEs bring open-source AI into real products and services.
• JAAI (Just Add AI GmbH)coreGTM or Marketing PartnerJust Add AI delivers custom AI solutions for enterprise clients across industries. With elluminate integrated into their delivery workflow, their team systematically evaluates every solution before client handoff, ensuring AI shipped is production-proven and reliable. CEO Enno Röhrig is a public reference customer.
• BITMARCKcoreChannel Partner/ Reseller/ DistributorGermany's largest health insurance IT provider runs elluminate on a dedicated enterprise deployment, evaluating and improving various AI use cases serving millions of insurance employees and insured members across their network of statutory health insurance companies.
• T-SystemscoreChannel Partner/ Reseller/ DistributorT-Systems, a subsidiary of Deutsche Telekom, is listed as a client and partner. Provides enterprise IT and digital transformation services that can integrate ellamind's AI agent evaluation and deployment platform for regulated enterprise customers.
• ITZBundcoreChannel Partner/ Reseller/ DistributorITZBund is the German Federal Government's IT service provider. Partnership enables deployment of ellamind's AI agent solutions across German public administration.
• hkk KrankenkassecoreGTM or Marketing PartnerGermany's largest statutory health insurance fund (GKV) with over 1 million insured members. Uses elluminate to continuously validate AI-powered website search for accuracy, medical advice rejection, and manipulation resistance. CEO Dr. Birger Schlünz is a public reference customer.

Scale indicators5 records

Recent moves6 records

Expansion highlights6 records

Broad incumbents

• Salesforce Agentforce: Enterprise platform for deploying AI agents with built-in governance, audit, and compliance. Competes in regulated enterprise agent deployment but at broader, less specialized scope.
• Datadog: Enterprise observability incumbent with LLM-specific monitoring, evaluation, and AI agent tracing capabilities. Stronger in large enterprise accounts but adjacent in functional overlap.
• Weights & Biases: Established MLOps and experiment tracking platform that has expanded into LLM evaluation, prompt tracing, and AI observability. Broader portfolio but overlapping developer-led buyer persona.

Direct peers

• Braintrust: Evaluation and observability platform for LLM applications with prompt management, scoring, and CI/CD-grade evaluation loops. Comparable in serving teams building AI agents into production.
• LangSmith (LangChain): LLM evaluation and observability platform tightly integrated with LangChain agents. Competes on evaluation, debugging, and monitoring workflows for production AI agents.
• Galileo (Galileo AI): AI evaluation and observability platform focused on hallucination detection, RAG evaluation, and agent reliability. Closely comparable in production-grade agent quality workflows.
• Confident AI (DeepEval): Open-source LLM evaluation framework (DeepEval) plus hosted platform for agentic evaluation, regression testing, and red-teaming. Direct overlap in criteria-based scoring and experiment workflows.
• Arize AI: AI observability and evaluation platform offering tracing, evaluation, drift detection, and experiment tracking for LLM and agent systems. Strong overlap in ELT-style AI quality workflows.
• Langfuse: Open-source LLM engineering platform with evaluation, tracing, prompt management, and observability. Closest direct competitor in agentic evaluation with overlapping model-agnostic positioning and PLG motion.

Emerging players

• WhyLabs: AI observability platform with LLM-specific evaluation, drift monitoring, and guardrails. Overlaps in real-time monitoring and policy enforcement for production AI.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• GitHubgithub.com/ellamind
• LinkedInlinkedin.com/company/ellamind

Compliance4 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles10 records

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about ellamind