Krisp

What Krisp does

Krisp builds real-time Voice AI for the real world, improving live voice conversations with noise cancellation, voice isolation, bidirectional accent conversion, speech translation in over 60 languages, agent assist, speech analytics, and infrastructure for AI voice agents. Trusted by businesses, developers, and millions of users worldwide, Krisp is deployed on 200M+ devices and processes 80B+ minutes of voice conversations every month. Built for real-world conditions, Krisp helps organizations make voice communication clearer, smarter, and easier to scale. Krisp's real-time Voice AI products improve voice productivity in communication through AI-powered technologies such as Noise Cancellation, Accent Conversion, Voice Translation, Transcription, Summarization and more. Krisp works across all audio hardware configurations and applications that support voice communication and has processed over 4T minutes of voice conversations so far.

Products

• Meeting AIAI for meetings for individuals and teams, offering in-meeting superpowers like noise-free calls, instant transcripts, and action items, and automating post-meeting tasks like summaries and CRM updates. It also helps prepare before meetings with AI meeting agents, smart agendas, and insights.
• Krisp for BPOs and Call CentersBoosts call center productivity with Voice AI, offering features like Accent Conversion, Voice Translation, Noise Cancellation, Agent Assist, AutoQA, Call Summary, Compliance Macros, Knowledge Chat, Call Scoring, Performance Monitoring, Compliance Monitoring, Deepfake Detection, Fraud Detection, and Agent Voice Verification.
• Krisp for Voice AI DevelopersProvides real-time Voice AI models and SDKs (VIVA SDK, RTC SDK) for cleaning, understanding, and transforming audio, as well as a self-serve Translation API. These tools are production-ready and built for various platforms including Windows, Mac, Linux, Web, iOS, and Android.
• VIVA SDKVoice infrastructure for voice AI agents, with lightweight, CPU-deployed models that operate on audio alone. Includes VIVA 2.0 for server-side voice isolation, Turn Prediction, Interruption Prediction, and Voice Activity Detection.
• RTC SDKEnhances the quality of communication in human-to-human calls and meetings. Includes models for accent conversion, background voice cancellation, and de-reverberation.
• Voice Translation APIReal-time speech-to-speech translation built for accuracy-critical applications, supporting 61 languages with any-to-any pair translation, custom vocabulary, and background voice cancellation.
• Speech AnalyticsAI Speech Analytics for Call Centers, providing automatic post-call scoring, compliance monitoring, and performance insights across every conversation. It includes Call Scoring, Performance Monitoring, and Compliance Monitoring features.
• Voice SecurityReal-time protection against synthetic callers, social engineering, and agent identity fraud for contact centers. Includes Deepfake Detection, Fraud Detection, and Agent Voice Verification.

Key capabilities

• AI noise cancellation
• AI accent conversion
• AI voice translation
• Transcription & Recording
• AI Meeting Notes
• Summaries
• CRM updates
• Centralized Action items
• AI meeting agent
• Smart agendas
• Recording presets
• On-time notifications
• Multilingual meeting support (16 languages for transcripts and summaries)
• 1-click sharing (Slack, timeline)
• In-person meeting support (mobile and desktop recording/transcription)
• Ask Krisp (AI chat for meeting content)
• Workspaces (organize meeting knowledge)
• Agent Noise Cancellation
• Customer Noise Cancellation
• Agent Voice Isolation
• After-call Summary
• SSO & SCIM
• Usage analytics
• Agent Accent Conversion (American and British English outputs)
• Customer Accent Conversion
• 60+ languages, any-to-any pair translation
• Industry vocabulary & dictionary
• AutoQA
• Background voice cancellation
• Echo and reverberation removal
• Bidirectional communication
• Call Summary
• Compliance Macros
• Knowledge Chat
• Call Scoring
• Performance Monitoring
• Compliance Monitoring
• Deepfake Detection
• Fraud Detection
• Agent Voice Verification
• Real-time Voice AI core
• Voice Governance
• VIVA SDK (Voice Isolation v3, Turn Prediction v3, Voice Activity Detection, Interruption Prediction, Signal Detectors)
• RTC SDK (Accent Conversion, Background Voice Cancellations, De-reverberation)
• Voice Translation API (61 languages, 96% accuracy, custom vocabulary, accent robust)
• Automated Call Scoring
• Customizable scoring criteria
• Structured scorecard evaluations
• Trends by agent, team, and time period
• Customizable compliance rules
• Real-time Deepfake Detection
• Real-time Fraud Detection
• Continuous Agent Voice Verification

Use cases

• Improving live voice conversations
• Noise cancellation in meetings and calls
• Voice isolation
• Bidirectional accent conversion
• Speech translation
• Agent assist in call centers
• Speech analytics for call centers
• Infrastructure for AI voice agents
• Streamlining note-taking in meetings
• Automating post-meeting tasks (transcripts, summaries, CRM updates)
• Preparing for meetings with AI agents and smart agendas
• Ensuring clear communication across global teams (multilingual support)
• Recording and transcribing in-person meetings
• Organizing and accessing knowledge from team conversations
• Boosting call center productivity
• Enhancing customer experience through clear communication
• Authorizing transactions and account changes securely
• Supporting operations across geographies and mixed working models
• Monitoring compliance in regulated environments (e.g., HIPAA)
• Protecting agents without CCaaS-specific integration
• Cleaning, understanding, and transforming audio for voice applications
• Server-side voice isolation for AI agents
• Predicting speaker turns and interruptions for AI agents
• Detecting voice activity robustly
• Converting agent accents to match customers in real-time
• Removing background noise and echo from agent and customer sides
• Real-time speech-to-speech translation for critical applications
• Automatic post-call scoring and performance insights for call centers
• Monitoring compliance on every call
• Detecting synthetic or cloned voices in real-time
• Identifying social engineering patterns during live calls
• Verifying agent identity continuously

AI approach

Krisp builds real-time Voice AI for various applications including noise cancellation, voice isolation, accent conversion, speech translation, agent assist, speech analytics, and infrastructure for AI voice agents. They develop their own models and SDKs for these functionalities, processing vast amounts of audio data to train and improve their AI.

Tech named: Voice AI, AI noise cancellation, AI accent conversion, AI Meeting Notes, AI meeting agent, AI Voice SDK, VIVA SDK, RTC SDK, Translation API, Voice Isolation, Turn Prediction, Voice Activity Detection, Interruption Prediction, Signal Detectors, Background Voice Cancellation, Outbound Noise Cancellation, Inbound Noise Cancellation, Speech-to-Speech Translation API, AI Speech Analytics, Deepfake Detection, Fraud Detection, Agent Voice Verification

Industries served

• Software Development
• Call Centers
• BPOs
• Financial services
• Healthcare
• Insurance

What it says sets it apart

• Real-time Voice AI for the real world
• Deployed on 200M+ devices and processes 80B+ minutes of voice conversations monthly
• Processes 4T+ minutes of voice conversations so far
• #1 noise cancellation
• Seamlessly integrates with all CCaaS and UCaaS platforms
• Real-time Voice AI core stays at the center; Voice Governance adds intelligence and security
• Sits between the agent’s headset and CCaaS/UCaaS stack, processing calls in real time with no telephony integration required
• 99% Increase in NPS, 25% Increase in FCR, 30% reduction in cost for call centers
• Supports 17+ dialects for accent conversion with more coming soon
• Supports 80+ languages for voice translation
• Voice Translation API has 96% accuracy on real calls, not just clean benchmarks
• Voice Translation API handles names, numbers, emails accurately
• Voice Translation API supports locale-specific variants and custom vocabulary/dictionary
• Voice Translation API includes built-in background voice cancellation and is accent robust
• Voice Translation API is self-serve with API key access
• VIVA SDK models are lightweight, CPU-deployed, and operate on audio alone
• VIVA 2.0 improves WER and includes multilingual Turn Prediction
• Voice Security runs inside existing Krisp deployment, CCaaS-agnostic, zero new infrastructure
• SOC 2 Certified
• GDPR Compliant
• HIPAA Compliant
• PCI-DSS Certified
• Enterprise-grade protections for data security and privacy
• Offers private transcription & recordings (on-device) for Enterprise plans
• Dedicated account manager for Enterprise plans
• Flexible pricing plans for individuals, teams, BPOs, call centers, and developers