Europe’s No. 1 network services provider
- ≈ €5bn
- Annual revenue (2025)
- ≈ 18,000
- Employees
- 10
- Country markets
- 1,800+
- Employees in Germany
AI agents that handle your calls, chats, bookings, and follow-ups; 24/7, automatically. Kitalk replaces legacy IVR with autonomous, action-oriented voice agents, built on our own State Graph technology, wired directly into your ERP, CRM and databases, and in production in days rather than quarters.
Circet, which builds and maintains Europe’s telecom infrastructure, runs internal and external customer service in Germany on Kilevel’s AI agents.
Kitalk carries the service line on its voice AI architecture, with no queue.
Kios agents complete each request inside Circet’s own systems.
Not a fixed connector catalogue: a builder. Describe the operation and the Kilevel assistant builds the whole agent from it: the state graph, the tools with their parameter schemas and execution plans, the API integrations behind them, the data it reads and writes back, the notifications it fires and the point where it hands over to a person. It assembles from your own library of reusable pieces where one fits, and writes new ones where none does.
Kilevel writes them in a sentence.
Autonomous discovery, real-time data sync, and bidirectional tool execution eliminate integration maintenance entirely.
Focus on what truly matters. AI takes care of the operational lift.
Autonomous discovery, real-time data sync, and bidirectional tool execution eliminate integration maintenance entirely.
Automatic MCP discovery
Kilevel automatically discovers, connects to, and digests MCP (Model Context Protocol) servers, reading their exposed tools, resources, and capabilities and making them instantly available across your flows, agents, and automations. Any CRM or ERP with an MCP endpoint is integrated without a single line of connector code.
Fully API-driven, in both directions
Everything is under granular control. Trigger calls and automations from your own tech stack, and retrieve the results of your operations at any time, from any CRM, ERP, or custom system.
Phone, messaging, email and website conversations on one operations surface, with AI analysis turning each one into a clear next action.
The same surface for five agents or a thousand, with per-tenant data separation, scoped API keys, roles for every operator and a complete audit trail behind it.
A visual workflow canvas, trigger, AI step, branch and act, across 35+ node types.
Every call, chat, and message is processed the moment it ends. Sentiment, summaries, and next steps are derived alongside custom signals and action triggers, so your team always has full context without manual review.
Future AI calls can be generated automatically for appointment confirmations, call-me-later requests, and any other follow-up workflow.
Every tool execution can be gated, deterministically, on conditions over state and arguments, or agentically, on a judgement about the conversation. A tool can stay visible and still have the call refused, with the reason handed back to the AI. Refunds over $500 without approval simply do not run.
An active software company since 2025. Its core product, Kitalk, replaces legacy IVR in professional call centres with autonomous, action-oriented voice agents.
No missed calls, no queue
Every inbound call is answered on the first ring, at any hour, at any volume.
Self-service configuration
Agents are built, tested and changed by your own team, in natural language.
Direct and indirect cost reduction
Fewer seats on repetitive traffic, and fewer escalations from mishandled calls.
In production with contact centres worldwide
Running live traffic today, not a pilot programme.
Complete autonomy: End-to-end AI management for outbound campaigns such as debt collection, as well as inbound customer service. The agent completes the work rather than just routing it.
Deep integration: A dynamic architecture that reads from and writes to your ERP, CRM and databases in real time, during the call, rather than reconciling afterwards.
Unlimited capacity: Concurrent calls scale with demand instead of with headcount, so queues do not form at peak and capacity is never the constraint on a campaign.
Data sovereignty: Optional on-premises deployment and dedicated management for maximum data security and the protection of sensitive customer information.
One vertically integrated stack, developed independently from the ground up. The same state graph, telephony layer, tool runtime and audit trail sit under all four, which is why an agent built for inbound service can be pointed at an outbound campaign without being rebuilt.
The shared substrateDeploy deterministic yet highly agentic voice and chat systems within minutes, built from the ground up on our proprietary State Graph technology.
Omnichannel interface: Continuous conversations over phone, SMS, email and WhatsApp.
AI-to-AI testing: A sandbox simulation suite to validate performance and stress-test before traffic.
Smart call forwarding: Seamless hand-off rules between AI and human agents.
Contact software designed for mass inbound and outbound strategic campaigns, operated by the team that owns the number.
Campaign engine: High-volume automated voice pipelines on custom schedules.
Human-in-the-loop: Intercept, monitor and guide live agentic calls as they run.
A/B testing: Split traffic between distinct AI models or human test groups.
Offload long-running analytical operations and background enterprise logic processing to agents that report back with structured output.
Document parsing: Deep text extraction and structured database formatting.
Dynamic CRM sync: Automatic connection construction using real-time API triggers.
Research pipelines: Long-running deep-dive workflows with structured reports.
A strategic communication bridge built explicitly for human and AI agents working the same queue.
Unified interplay: A live agent transfer desk with rich contextual background transfer.
Real-time copilot: On-screen transcription, live CRM inserts and next-action prompts.
Omni-inbox: One interface managing SMS, WhatsApp, email and live calls.
Over a decade of deep engineering research in AI tech and voice processing, building proprietary models that set state-of-the-art standards.
Our hyper-scalable cloud infrastructure is engineered to handle thousands of concurrent voice calls per user, running fully autonomously.
High-precision conversational accuracy trained on millions of real dialogue interactions to understand complex intents.
Deploying active operator commands and isolated pipelines results in up to 40% higher customer engagement compared to legacy systems.
The platform is built, run and supported by the engineers who wrote it. Isolated tenant data stores, field-level access control and a node-by-node audit trail are product behaviour rather than roadmap items, and security, data-protection and procurement review are part of onboarding rather than a later conversation.
Anyone can show a convincing call. Running agents across a real organisation asks harder questions: who may do what, what happens when a model is wrong, and what the regulator sees. All of it is platform configuration, not a services engagement.
Kilevel is AI-first in its own architecture, not only in what it sells. The assistants have complete control over the platform surface, so everything on this list is built and changed by describing it, rather than configured by hand or waited for in a release.
Less hand-built surface means less to keep working: a lower cost to run over the life of the deployment, and a change that lands the day it is asked for rather than in the next quarter.
Managers, employees, call-centre agents, task agents and integrators each see their own slice. API keys carry per-endpoint permissions rather than full account access, and two-factor sign-in protects the console.
Run client workspaces with their own branding, data stores, sub-users and campaigns. Agencies delegate work into a tenant under an explicit permission, with its own budget and its own billing attribution.
Every tool call can be gated, deterministically on state and arguments, or agentically on a judgement about the conversation. A refused call hands the reason back to the model instead of silently proceeding.
Recording consent as a node with a deterministic keyword check, an AI-disclosure opener that fires before anything else, and sensitive stretches bracketed out of the recording. Call recordings are stored in the EU.
Concurrency ceilings per number, per line and per user keep one campaign from starving the rest. Work agents carry spend, token, model-call and tool-call caps, for themselves and across every sub-agent they spawn.
Admin logs and app events record who changed what. Every run keeps its node-by-node trace, flows are versioned and restorable, and data transfers out on request. Trigger anything over REST and receive events by webhook.
Nightly AI-to-AI regression suites grade your agents against real scenarios and alert the team when a score slips, so a change in behaviour is caught by the platform, not by a customer.
Ready-to-deploy agents for every sector and workflow, as a template your team adapts, not a finished product you have to accept as shipped.
Orchestrate specific customer interaction pipelines natively.
Highly customized templates for professional sectors.
Everything an operation runs on, in one system: the assistant that builds your agents, the channels and campaigns that reach people, the records and calendars they read and write, and the control room, analytics and compliance footing behind them. Eleven capabilities on one runtime, with nothing to integrate in between.
The world's best AI flow generation framework. Fully AI-assisted, beyond simple support bots. Describe integrations in one sentence; our assistant builds the connection in seconds.
Every building block is native and reusable across Flows, Agents, Notification Rules, AI Analysis, Reporting, and more.
Launch high-performance automated outbound campaigns that scale seamlessly, blending AI precision with human lead-generation strategies.
“Customer asked about setup — mention out-of-the-box Slack notifications.”
Configure rules to automatically schedule and trigger follow-up AI calls based on recipient answers, customer events or CRM states.
Embed voice, chat, hybrid callbacks, or animated human-like avatars on your website. Features instant speech-to-text and automated text-to-data parsing.
Manage your client base, interactions, services, feedback forms, and customer histories natively without external dependencies.
AI agents browse custom calendars, book appointments directly, and trigger SMS or email notifications automatically to minimize no-shows.
A powerful drag-and-drop designer for operators next to robust REST APIs, webhooks, and tool definitions for custom system integrations.
POST /v1/calls
{
"agent": "Sarah",
"phone": "+491609..."
}Monitor active concurrency, review live whispers, issue passive operator commands, and scale lines dynamically in real-time.
Allow immediate human intervention. Escalate automated calls to live support teams mid-conversation without interrupting the customer flow.
Gain profound insights from every call with pre-configured prompts that automatically score lead quality and trigger CRM webhooks.
Platform, database and recordings hosted in Frankfurt, Germany, with telephony infrastructure in a German data centre. Isolated database deployments available on request.
Autonomous speed with human oversight. Built for high-stakes enterprise workflows.
Pair autonomous AI agents with human intelligence. Run high-volume campaigns where AI qualifies intent and hands off to specialists, or construct multi-step workflows requiring instant human authorization before critical actions execute.
Blended campaigns & live handoff
Dispatch campaigns where AI agents perform high-volume outreach, qualify customer intent, and handle discovery—then escalate complex conversations to human agents in real time with complete dialogue history, sentiment signals, and live guidance.
Instant approval gates & guardrails
Design end-to-end workflows in seconds where autonomous agents prepare decisions but must pause for human sign-off before binding commitments, CRM edits, or financial transactions take effect.
| Compared against | Dimension | The alternative | Kilevel | |
|---|---|---|---|---|
| Legacy call centres / IVR | Conversation model | A fixed decision tree. Anything outside it becomes a transfer. | A state graph: deterministic where the process demands it, agentic where the caller does. | |
| Changing behaviour | A change request, a vendor ticket, a release window. | Described in natural language by your own team, versioned, and reversible. | ||
| Systems of record | Middleware and nightly reconciliation. | Read and write during the call, over MCP, REST or a generated tool. | ||
| Capacity | Bounded by licensed seats and trunk provisioning. | Scales with demand; concurrency ceilings are a policy you set, not a hard limit. | ||
| Voice AI platforms | Behaviour you can state | Prompt-driven. What the agent will do can only be described after it has done it. | A deterministic state graph bounds every step; the generative layer decides freely only inside it. | |
| Release testing | A person with a headset before go-live, covering what they remembered. | AI-written test calls run against every release as an automatic gate, with your acceptance criteria. | ||
| After the call | It speaks, then stops. A person types the outcome in. | Fields, score and transcript written back, follow-up triggered, callback booked, in the same run. | ||
| Channels | Voice. Chat and messaging are a different product, with a different agent. | Phone, WhatsApp, SMS, email and web on one agent, one inbox and one history. | ||
| Running an operation | An API and a dashboard. | A contact centre: campaigns, queues, softphone, live whisper, barge-in and warm hand-off. | ||
| Agent studios | Who it was built for | Developers. Shared tenants, coarse permissions, thin audit trails. | The people who sign for it. Isolated tenants, role- and field-level access, retained decision logs. | |
| Telephony | Bring your own carrier, then build the contact centre around it yourself. | Carrier-grade SIP and PSTN, numbers, trunks and routing inside the platform. | ||
| Changing an agent | A ticket to a certified consultant, billed by the hour, delivered in weeks. | Described in natural language, implemented by the platform's own AI, then tested automatically. | ||
| Data and residency | Wherever the vendor hosts it. | Isolated per-tenant stores, field-level visibility, EU recording storage, on-premises optional. | ||
| Assurance | Manual sampling after the fact. | AI-to-AI suites, AI-judged reply quality and randomised A/B on live traffic. |
Every alternative above is one piece of this. Matching it means an IVR replacement, a voice vendor, an agent framework, a contact-centre suite and the integration work between them, each with its own contract, its own data model and its own audit trail. Here it is one platform, one tenant, one record of what happened. The seams between those products are the part nobody sells you, and the part you end up owning.
The fastest way to evaluate this is not a slide deck. Pick a single inbound queue or one outbound campaign, run it on Kilevel alongside your current setup, and compare answer rate, handling time and completion against the estate you already trust.
Procurement, security review and on-premises questions go straight to the team that built the platform.
Everything you need to know about Kilevel.
Kilevel makes three AI products on one platform. Ki is the personal AI executive assistant for professionals and teams: it gets to know your world, keeps track of your work across days, follows up, keeps your team in sync, thinks ahead even while you sleep, and levels up the longer you work together, on WhatsApp, phone and email and inside ChatGPT, Claude, Gemini and Grok. Kitalk is the AI contact centre for businesses: voice agents for inbound and outbound calls, WhatsApp, SMS, email, web widgets and hybrid human queues. Kios is the next-level AI agent operating system both run on: State Graph agents, automations, background Work agents and agent databases.
Legacy IVRs follow rigid scripts. Chatbots handle text in isolation. Kilevel is an AI-native operations platform that handles real phone calls, WhatsApp, SMS, email, and web conversations simultaneously, with genuine contextual understanding. Agents adapt mid-conversation, book appointments, trigger follow-up automations, analyze every interaction, and hand off to humans when appropriate. The entire stack runs from a single platform: inbound queue, outbound campaigns, post-call intelligence, and operational reporting.
Yes. You can build and test your AI agent for free using our library of ready-made industry templates or by starting from scratch. Design call flows, configure channels, and connect integrations before going live. When you deploy, you pay a monthly subscription fee, which already includes a bundle of free AI minutes per month. Any usage beyond the included allowance is billed on a pay-as-you-go basis. No seats, no surprise fees.
No seats, no per-user charges. Kilevel combines a monthly subscription with usage-based pricing: each plan includes a pool of free AI minutes per month, and additional usage beyond that is billed at a per-minute rate. Building, testing, and configuring agents is free. Once live, costs scale predictably with your volume. Enterprise customers with high-volume commitments can discuss custom pricing agreements.
Most teams ship their first agent within a day. Kilevel comes with a library of industry-specific templates that cover the most common call flows and channel configurations out of the box. Pick a template, configure your agent's persona and rules, connect your tools, and deploy, no code required. Custom flows are built visually, and the AI flow assistant can generate an entire flow structure from a single descriptive sentence.
Kilevel gives your team leverage, not a replacement. AI agents absorb the volume of after-hours calls, common questions, appointment bookings and multilingual conversations, so your staff can focus on complex cases that genuinely need a human. Built-in sentiment analysis and configurable handoff triggers mean the AI always knows when to step aside.
The platform, the database and your recordings run on Google Cloud in Frankfurt, Germany, and the telephony infrastructure sits in a German data centre. Speech and language providers are used through their EU endpoints wherever the provider offers one. Where a provider processes outside the EU, that transfer runs on Standard Contractual Clauses, and every such provider is named with its processing region in our public Trust Center. An Article 28 data processing agreement applies from day one, with no separate negotiation, and we never use your data to train models.
Yes. Agents can detect a caller's language and respond accordingly, or you can configure language-specific flows per channel. This is particularly relevant for European businesses serving customers in German, French, Spanish, Dutch, and other languages alongside English, all within the same platform, without duplicating your entire agent setup.
No. The platform is designed for operators first. Flow building, agent configuration, channel setup, and integration connections are all no-code. Developers who want deeper control get a full REST API, webhooks, MCP-compatible tool discovery, and prompt-level customization, but none of it is required to run a production deployment. Non-technical teams operate Kilevel every day.
Completely. Every flow, agent behavior, escalation rule, and analysis signal is configurable, visually, without code. You define the tone, language, fallback responses, and handoff conditions. Advanced users can go further with prompt modules, templating parameters, flow templates, and a full REST API. If your workflow is unusual, Kilevel almost certainly supports it.
Complete control. You define tone, language, permitted topics, and response templates using prompt modules and templating parameters, reusable building blocks your agents draw from consistently. Escalation rules trigger human handoff based on sentiment, topic, or conversation outcome. Every configuration is fully testable in a sandbox before it touches a real customer.
You have full control over what your agent says. Define tone, language, and exact response templates. Escalation rules let you specify when a conversation gets transferred to a human, based on topic, negative sentiment, or repeated failed attempts. Everything is testable before it goes live, and you can iterate on agent behavior at any time.
Yes, and often without any setup at all. Kilevel supports MCP (Model Context Protocol), which means it can automatically discover and digest integrations from modern CRMs, ERPs, and data platforms. For everything else, there are webhooks, a REST API, and pre-built connectors. Your data flows in and out on your terms.
Both directions are fully supported. Kilevel uses MCP (Model Context Protocol) to automatically discover and integrate modern CRM and ERP systems, often with zero connector setup. In the other direction, Kilevel's REST API lets your own tech stack trigger calls, campaigns, and automations externally, and pull conversation results, analysis, and outcomes back into your systems on demand.
Yes, automatically. Kilevel scales concurrent lines dynamically, whether you're handling 10 conversations or 10,000, the infrastructure adjusts without manual intervention. The Operations Control Center gives you live visibility into concurrency, queue depth, and agent performance, so you're never operating blind at scale.
Your data sits in Frankfurt, Germany: platform, database and recordings on Google Cloud, telephony in a German data centre. Credentials are encrypted at the column level with AES-256-GCM, every query is scoped to its owning account, sub-users carry project-specific roles, and administrative actions are written to a reviewable audit trail. We process only on your documented instructions and never train models on your data. The full subprocessor list, with the processing region of each, is published in our Trust Center.
Kilevel agents are designed to manage uncertainty without frustrating the caller. They can ask clarifying questions, offer alternate paths, or escalate to a live operator, automatically, based on rules you set. The Attention Suite enables immediate human takeover mid-conversation without any disruption to the customer experience.
Kilevel agents handle uncertainty without derailing the experience. They can ask clarifying questions, reroute the conversation along alternate paths, or escalate to a live operator, all automatically, based on configurable rules. The Attention Suite lets operators intervene instantly mid-call, taking over the conversation without the customer noticing any disruption.