# Cognigy — AI Agents & Assistants

Cognigy by Cognigy GmbH holds a panel rating of 2.5/5 from 6 AI judges (category median 3/5), as of 2026-10-28. Provenance: EU origin, foreign-owned; sovereignty: not determined.

- Vendor: Cognigy GmbH (https://www.cognigy.com)
- Category: [AI Agents & Assistants](https://whats-best.ai/software/ai-agents.md)
- Provenance (editorial classification): EU origin, foreign-owned
- Canonical page: https://whats-best.ai/software/ai-agents/cognigy
- Alternatives: https://whats-best.ai/software/ai-agents/cognigy/alternatives ([markdown](https://whats-best.ai/software/ai-agents/cognigy/alternatives.md))
- Data as of: 2026-10-29 · evaluated 2026-10-28 · oldest source captured 2026-10-26

**AI-generated.** Every score, rationale and verdict here is written by a panel of AI judges from public evidence — an analysis, not a hands-on test. Every fact quotes the vendor page it was read from. No score can be bought. How scores are made: https://whats-best.ai/methodology

## At a glance

- Vendor: Cognigy GmbH
- Headquarters: Not stated on the vendor’s pages as of 26 Oct 2026
- Data hosting: Not stated on the vendor’s pages as of 26 Oct 2026
- Free plan: Not stated on the vendor’s pages as of 26 Oct 2026

## The panel's verdict

Cognigy is an AI-agent platform for contact centres on voice and chat from Cognigy GmbH of Düsseldorf. Scoring is lopsided: integrations and knowledge and agent capabilities lead at 7-8, covering prebuilt connectors to CRM, ticketing, payments and 100+ more systems, full MCP client support with per-node tool filtering, and a Simulator suite running synthetic customers with success criteria, regression testing and version comparison. Data use and training is the weakest scored area outside pricing at 2-3: the published privacy policy governs the vendor's websites and its own support bot, and judges found no public information on platform retention or a data processing agreement; the spread reflects different weight on the website retention specifics. Processing and models scores sit at 4 with no variation — model vendors are named and selectable, but we found no public information on where inference runs. Sovereignty scores read 3-4: the contracting entity is German, the site brands the product NiCE Cognigy, and we found no public statement on where customer data is hosted. Pricing transparency reads 0-1, with only a free trial and a demo.

**Speaks for it:**
- Prebuilt connectors to CRM, ticketing, payments and 100+ more systems, with a documented REST API and developer CLI
- Full MCP client support with automatic schema updates and per-node tool allow- and deny-lists
- Simulator evaluation suite with synthetic customers, scheduled and regression testing, automatic scoring with full transcripts, and version comparison
- Model orchestration across OpenAI, Azure OpenAI, Anthropic, Google and Aleph Alpha with the LLM of your choice
- Contracting entity and GDPR controller is Cognigy GmbH of Düsseldorf with the North Rhine-Westphalia authority as regulator

**Held against it:**
- The published privacy policy governs the vendor's websites and its own support bot, not the deployed platform
- We found no public information on where inference runs, EU processing options or bring-your-own-key deployment
- We found no public statement on where customer data is hosted; the subprocessors listed cover website tools, with Standard Contractual Clauses stated only for HubSpot
- We found no public information on SCIM provisioning, audit logs or group-level permissions over models, agents and connectors
- We found no public information on human approval before consequential agent actions or on production triggers and schedules

**Best for:**
- You run a contact centre on voice and chat and need prebuilt connections into CRM, ticketing and payments systems
- Your team must prove agent quality before release through synthetic customer simulations, regression runs and version comparison
- You want to orchestrate models from several vendors, including the European option Aleph Alpha, through one interface
- You need agents portable across development, staging and production with tool allow- and deny-lists at each node

**Avoid if:**
- You must review platform-grade data processing terms and retention commitments before signing — the published privacy policy covers the vendor's websites and its own support bot
- You must pin inference and hosting to the EU — the pages name providers including OpenAI, Anthropic and Google, and we found no public statement on where inference or hosting run
- You plan agents running unattended in production systems — the vendor states that MCP authentication and permission handling are still evolving
- You need published prices to budget — the only entry points shown are a free trial and a demo request

## Scores by judge and criterion

Scores out of 5, each judge scoring independently against the published rubric (v1). "n/r" = not rated: the vendor publishes no price, or the judge found no public information on that criterion.

| Criterion | The Head of IT | The Data Protection Officer | The Automation Lead | The Customer-Service Director | The Works Council Member | The Skeptic |
| --- | --- | --- | --- | --- | --- | --- |
| Data use, training & retention | 1/5 | 1/5 | 1/5 | 1.5/5 | 1.5/5 | 1/5 |
| Models & processing location | 2/5 | 2/5 | 2/5 | 2/5 | 2/5 | 2/5 |
| Admin, identity & access controls | 2/5 | 2.5/5 | 2.5/5 | 2/5 | 2.5/5 | 2/5 |
| Integrations, connectors & knowledge | 3.5/5 | 3.5/5 | 4/5 | 4/5 | 3.5/5 | 4/5 |
| Agent capabilities & guardrails | 3.5/5 | 4/5 | 3.5/5 | 4/5 | 3.5/5 | 3.5/5 |
| European sovereignty | 2/5 | 2/5 | 1.5/5 | 2/5 | 2/5 | 2/5 |
| Pricing transparency | n/r | n/r | n/r | n/r | n/r | n/r |
| **Weighted verdict** | **2.5/5** | **2/5** | **3/5** | **3/5** | **2.5/5** | **2.5/5** |

### The judges

- **The Head of IT** (v1.0): Has to roll an AI assistant out to eight hundred staff without a second shadow-IT wave. Wants SSO and SCIM, groups that decide who gets which model, agent and connector, answers that respect SharePoint permissions, audit logs for the SIEM, and a deployment option that is not "trust us".
- **The Data Protection Officer** (v1.0): Signs the DPA and gets asked where a prompt goes. Wants every model provider named as a subprocessor with its region, training excluded for all of them, retention the company controls, and an EU-only switch that also covers agents and connectors — not just the chat window.
- **The Automation Lead** (v1.0): Builds the agents that file tickets, draft offers and chase approvals across the ERP, CRM and wiki. Wants tools and actions with scoped permissions, MCP and an API, approval steps before anything is sent, run logs, and a way to test an agent before the whole company uses it.
- **The Customer-Service Director** (v1.0): Runs a contact centre with forty agents and wants half the routine calls and chats answered by an AI agent. Wants handover to a human with the full context, actions in the order and ticket systems, evaluation before go-live, and a price per conversation the CFO can forecast.
- **The Works Council Member** (v1.0): Has a co-determination right the moment a tool can monitor staff, and an AI assistant logs every prompt. Wants usage reports that can be aggregated or pseudonymised, prompt logging rules set out in writing, a human approving what agents do to colleagues, and no training on staff input.
- **The Skeptic** (v1.0): Has read "European AI" on a wrapper that sends every prompt to a US model, "agents" that are saved prompts, and "we never train on your data" with no word on the model provider behind it. Asks which model answers, where it runs, what it may do unsupervised, and what a thousand seats cost.

## European sovereignty — proven facts

Built only from facts on the vendor's own pages; a dimension we could not prove is left open, never scored as zero. Headline: not determined.

| Dimension | Finding | Points | Source |
|---|---|---|---|
| Legal entity | Not determined | — | — |
| Ownership | Not determined | — | — |
| Data residency | Not determined | — | — |
| Subprocessors | Not determined | — | — |

## Key facts

Each fact is the vendor's own statement, cited to the source page it was read from ([S*n*] → Sources).

**Compliance**
- Cookie consent: Remove cookies button on privacy policy page · yes [S4]
- Forrester leader: vendor claim on page · yes [S11]
- Ga ip anonymization: yes [S4]
- GDPR reference: yes [S4]
- Mcp standard maturity: MCP authentication and permission handling still evolving; Cognigy to support improvements in future releases [S9]
- Order processing: yes [S4]
- Third party transfers: yes · EU Standard Contractual Clauses [S4]

**Data**
- Automated decision making: no [S4]
- Ga retention: 14 [S4]
- Log ip retention: 1 week [S4]
- Marketing double optin: yes · 24 [S4]
- Marketing partner sharing: yes [S4]
- Support AI retention: 30 [S4]
- Third party sharing: consent, legal obligation, agreed purposes · no [S4]
- Trial data retention: 3 [S4]

**Integrations**
- Cli: Full-featured developer CLI integrates with CI/CD and source control [S2]
- CRM switching: MCP connector can be swapped on CRM migration without changing agents [S9]
- Extensions: Extension framework with hundreds of free plug-ins in a public repository [S2]
- Llm orchestration: centralized interface to orchestrate dozens of LLMs [S6]
- Mcp node: Tool Nodes · yes [S9]
- Mcp schema updates: yes [S9]
- Mcp setup: enter MCP server endpoint, no manual API call definition [S9]
- Mcp support: yes [S9]
- Mcp tool filtering: yes [S9]
- Nexus engine: Cognigy Nexus Engine [S9]
- No custom code: no [S9]
- Prebuilt tools: 100+ systems including CRM, ticketing, payments [S2]
- Rest API: Documented, secured REST API with full platform coverage [S2]

**Legal**
- Applicable websites: www.cognigy.com · hello.cognigy.com [S4]
- Consumer arbitration: Not willing or obliged to participate in consumer arbitration board proceedings [S3]
- Controller: Cognigy GmbH · Kesselstraße 3, 40221 Düsseldorf, Germany · Germany [S4]
- Dispute resolution: References EU ODR platform [S3]
- Entity name: Cognigy GmbH · Kesselstraße 3, 40221 Düsseldorf · Düsseldorf · HRB 79167 [S3]
- Forrester recognition: Named a Leader in Forrester Wave 2026: Conversational AI Platforms for Customer Service [S3]
- Lei: 9845002IFEDAF12DC714 [S3]
- Nice branding: Consent form references both NiCE and NiCE Cognigy [S6]
- Nice ownership: branding references both NiCE and NiCE Cognigy in privacy/consent context [S11]
- Regulator: North Rhine-Westphalia District Data Protection and Freedom of Information Officer (LDI NRW) [S4]
- Rights: access, rectification, deletion, restriction, objection, data portability per GDPR Art. 15-21 [S4]
- VAT id: DE309073054 [S3]

**Product**
- Adaptive content: yes [S2]
- Agent builder: Create AI agents with natural-language prompts [S2]
- Agent copilot: Agent Copilot [S3]
- Agent knowledge search: Intelligent knowledge search for human agents in real time [S8]
- Agentic AI: Agentic AI [S3]
- AI agents: AI Agents [S3]
- AI scenario generation: yes [S11]
- Analytics capabilities: analytics capabilities listed: intent and volume analytics, containment and completion metrics, drop-off and funnel insights, channel and handoff reporting, optimization and trend analysis [S7]
- API mocking: yes [S11]
- Architecture: microservices-based [S2]
- Automated qa: yes [S2]
- Automated qa: Automated generative Q&A using vector search and LLMs across knowledge sources [S8]
- Automatic scoring: Automatic scoring against configurable criteria with drill-down into failed conversations [S10]
- Built in ide: Write code inside the Cognigy.AI Flow Editor [S2]
- Channels: webchat · voice · SMS · messaging apps · yes [S2]
- Chat messaging: Chat & Messaging [S3]
- Containment claim: 20x higher containment within one month of go-live [S2]
- Context memory: Context-aware conversations; agents remember prior conversation [S8]
- CRM personalization: CRM integration personalizes answers per customer [S8]
- Cross lingual representation: Multilingual models build cross-lingual representations where the same concept maps to the same internal representation across languages [S5]
- Custom llm: yes [S8]
- Customer case: Lufthansa · 50 [S6]
- Dashboards: overview dashboard with success rate trend over time and recent simulation results [S11]
- Demo available: yes [S10]
- Digression handling: yes [S6]
- Empathetic responses: yes [S6]
- Entity extraction: yes [S6]
- Environment portability: yes [S2]
- Evaluation runs: Simulations run on demand, on schedule, or as automated regression testing [S10]
- Forrester leader: Leader in Forrester Wave 2026: Conversational AI Platforms for Customer Service [S1]
- Forrester leader: Leader in Forrester Wave 2026 Conversational AI Platforms for Customer Service, highest in Strategy category [S8]
- Free trial: Cognigy.AI trial · yes [S4]
- Input rephrasing: yes [S6]
- Knowledge AI: Knowledge AI [S3]
- Knowledge AI rag: Knowledge AI combines RAG retrieval with task automation [S6]
- Language autodetect: LLM-based automatic detection of the user's spoken language [S6]
- Leader recognition: Named a Leader in the Forrester Wave™ 2026: Conversational AI Platforms for Customer Service [S7]
- Live debugging: yes [S2]
- Llm assisted design: Generate agent resources with LLMs; visual process orchestration [S2]
- Llm intent refine: GenAI embedding model intent optimization plus LLM NLU fallback [S6]
- Llm variation runs: each run introduces LLM-generated variation; run count is team-configurable [S11]
- Localization obsolete longtail: Universal language model renders localization obsolete for the long tail of less frequently used languages [S5]
- Mcp low code flows: MCP Node usable in flows for live conversation data via configuration [S9]
- Model agnostic: model-agnostic, integration-friendly approach [S9]
- Monolingual model tradeoffs: Monolingual models are described as more optimized, faster, and more performant [S5]
- Multilingual consistency: Multilingual consistency across languages and regions is measured [S10]
- Multilingual model recommendation: yes [S5]
- Multilingual training: yes [S6]
- No data provision required: no [S4]
- No intent training: Deploy via document upload without intent training or complex flows [S8]
- Outcome claims: pre-deployment stress testing · automated simulations replacing manual QA · enterprise-grade reliability validation · holistic technical and CX evaluation [S11]
- Output format control: Configurable UX: plain text or rich media, with or without sources [S8]
- Platform modules: Cognigy.AI Overview, Agentic AI, Knowledge AI, Voice Connectivity [S7]
- Prioritized routing: yes [S6]
- Rag answers: Knowledge AI combines conversational AI, LLMs and vector search to answer customer questions from enterprise knowledge in real time [S8]
- Realtime preview: yes [S2]
- Regression tracking: Success rates tracked over time to catch regressions [S10]
- Response traceability: yes [S8]
- Run transcripts: yes · yes [S11]
- Scalability: Thousands of conversations with zero downtime [S2]
- Scenario elements: 4 · persona · mission · success criteria · maximum conversation turns [S11]
- Scenario generation: Scenarios can be written manually or generated from existing AI agents and real transcripts [S10]
- Security features: SSO · RBAC · end-to-end encryption · fine-grained data controls [S2]
- Separate locale intent management: Cognigy supports intent editing and training per separate locale [S5]
- Simulation scope: simulation runs a single scenario against a specific Agent version, Flow, and Locale [S11]
- Simulation triggers: yes · yes [S11]
- Simulator purpose: Simulator stress-tests AI agents across thousands of realistic conversations against explicit success criteria [S10]
- Simulator purpose: Simulator is an agent evaluation suite for validating and optimizing AI Agents [S11]
- Simulator scenarios: Test scenarios use synthetic customers with persona, mission and success criteria [S10]
- Solutions: AI Agents, Phone & Voice, Chat & Messaging, Agent Copilot, Contact Center Integrations [S7]
- Source filtering: Data transparency: categorize sources and filter search scope [S8]
- Support AI agent: Cognigy Support AI Agent processes user input to answer queries [S4]
- Synthetic customers: synthetic customers as digital twins mirroring demographics, behaviors, language, intent [S11]
- Third party mocking: Simulator mocks third-party API responses (success, degradation, error) without touching live systems [S10]
- Transactions: Backend integrations for transactional queries end-to-end [S8]
- Transformer architecture: Universal model is based on Transformer model architecture [S5]
- Trial phone bot: demo phone bot processes call content, call date/time and caller phone number [S4]
- Universal language model: Universal Language model maps inputs in other languages (e.g. German, Japanese, Russian) to intents trained with English example sentences [S5]
- Version comparison: version comparison · multilingual consistency validation · change impact prediction [S11]
- Visual builder: One visual workspace to build, test and ship AI agents; no coding required [S2]
- Voice and chat agents: yes [S2]
- Voice chat messaging copilot: voice · chat · messaging · yes [S1]
- Voice connectivity: Voice Connectivity [S3]

**Subprocessors**
- Azure openai: Azure OpenAI (Microsoft) · answer Support AI Agent queries · no [S4]
- Cloudflare: Cloudflare · CDN and security for HubSpot-hosted sites [S4]
- Dreamdata: DreamData · B2B revenue attribution analytics [S4]
- Google analytics: Google Analytics · web analytics [S4]
- Microsoft clarity: Microsoft Clarity · behavioral metrics, heatmaps, session replay [S4]
- Youtube: YouTube (Google Ireland) · embedded video content [S4]
- Zendesk: Zendesk · customer support communication [S4]
- Zoominfo: ZoomInfo · business people and company profiles [S4]

**Support**
- AI agent channels: AI agent · Zendesk · yes [S4]
- Documentation: yes [S3]

## Questions

### Is Cognigy European?

Partly. whats-best.ai classifies Cognigy as EU origin, foreign-owned.

### Where does Cognigy store data?

Not stated on the vendor’s pages as of 26 Oct 2026.

### Does Cognigy have a free plan?

Not stated on the vendor’s pages as of 26 Oct 2026.

### What are the best alternatives to Cognigy?

As of 28 Oct 2026, the panel rates [Langdock (4/5)](https://whats-best.ai/software/ai-agents/langdock), [Dust (3.5/5)](https://whats-best.ai/software/ai-agents/dust) and [Unique AI (3.5/5)](https://whats-best.ai/software/ai-agents/unique-ai) highest among the other 15 AI Agents & Assistants products evaluated. Ratings that look equal are rounded to half stars; the order uses the unrounded rating, then the name. The highest-rated European alternative is [Langdock](https://whats-best.ai/software/ai-agents/langdock) (EU-Made). Ratings are the AI panel’s opinion, formed from public evidence: [how scores are made](https://whats-best.ai/methodology#ordering). The full list, with European products first: [Cognigy alternatives](https://whats-best.ai/software/ai-agents/cognigy/alternatives).

## What this page leaves out

- We found no public information on pricing on the pages we read (cognigy.com, cognigy.com/platform/ai-agent-studio, cognigy.com/legal-notice, cognigy.com/privacy-policy, cognigy.com/product-updates/cognigys-universal-language-understading-model, cognigy.com/product-updates/llm-conversational-ai and 5 more). If the vendor publishes it somewhere else, send us the page.
- 27 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material.
- 4 subprocessors facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material.
- 1 compliance fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 1 integrations fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 1 legal fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 1 support fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material.
- 1 sovereignty dimension could not be confirmed on the vendor’s own pages and is shown as unknown.

## Sources

- [S1] Vendor homepage: https://www.cognigy.com/ (checked 2026-10-26)
- [S2] Features page (AI Agent Studio): https://www.cognigy.com/platform/ai-agent-studio (checked 2026-10-26)
- [S3] Legal notice: https://www.cognigy.com/legal-notice (checked 2026-10-26)
- [S4] Privacy policy: https://www.cognigy.com/privacy-policy (checked 2026-10-26)
- [S5] Models & processing location — found from sitemap: https://www.cognigy.com/product-updates/cognigys-universal-language-understading-model (checked 2026-10-26)
- [S6] Models & processing location — found from sitemap: https://www.cognigy.com/product-updates/llm-conversational-ai (checked 2026-10-26)
- [S7] Admin, identity & access controls — found from sitemap: https://www.cognigy.com/faq/what-analytics-are-available-in-conversational-ai-platforms (checked 2026-10-26)
- [S8] Integrations, connectors & knowledge — found from sitemap: https://www.cognigy.com/platform/cognigy-knowledge-ai (checked 2026-10-26)
- [S9] Integrations, connectors & knowledge — found from sitemap: https://www.cognigy.com/product-updates/model-context-protocol-mcp (checked 2026-10-26)
- [S10] Agent capabilities & guardrails — found from sitemap: https://www.cognigy.com/platform/ai-agent-evaluation (checked 2026-10-26)
- [S11] Agent capabilities & guardrails — found from sitemap: https://www.cognigy.com/product-updates/meet-simulator-ai-agent-evaluation (checked 2026-10-26)

---

Cite as: whats-best.ai, "Cognigy — AI Agents & Assistants", AI panel analysis, as of 2026-10-29. https://whats-best.ai/software/ai-agents/cognigy

Corrections: https://whats-best.ai/correct/1033?target=page