Konvo

The agent writes into the merchant's commerce stack through a named skill catalog. Order management covers get order status, cancel order, edit order shipping address and refund order. Manage subscription deliveries and edit subscription shipping address sit under subscriptions, and close conversation, escalate to an agent and start a side conversation are general actions.

Connected systems include Shopify and WooCommerce for shop platforms, Mintsoft, Amphora and Sendcloud for fulfillment and logistics, Gorgias, Zendesk, Front and Freshdesk for helpdesks, and Klaviyo and Salesforce for marketing and CRM. Konvo sums this up as connecting a merchant's 3PL, shop system and CRM. Beyond structured connectors, the agent takes on tedious actions like calling a courier or emailing a provider, so it reaches parties with no integration at all. Konvo publishes no connector directory or count beyond that named set.

Sourcekonvoai.com AI platform section and FAQread 2026-09-05

Processes is one of five build components, alongside Knowledge, Persona, Skills and Recommendations. Konvo's agents act, reason and behave like a human, following complex processes and executing actions such as order changes, carrier communication and internal processes while holding the brand's tone. A cancellation, refund or update is resolved by reading the shop system and writing back to it. A courier incident is handled by calling a courier or emailing a provider, so the run continues through a third party before it can close.

A subscription is managed by adjusting deliveries. Skills compose into those processes and include branching outcomes, since escalate to an agent, close conversation and start a side conversation are terminal or forking steps the agent chooses between. The agent does the work instead of assisting with it, unlike helpdesks that add AI as a feature. Konvo names no runtime, step model, retry or failure handling.

Sourcekonvoai.com AI platform section and FAQread 2026-09-05

The agent is grounded on an indexed corpus with its own management surface, the Knowledge hub. It is one of five build components alongside Persona, Skills, Processes and Recommendations. One example hub holds 49 sources with a search box across them.

They include a crawled storefront at 23 pages, a crawled FAQ site at 5 pages, an uploaded document of order management protocols at 355 KB and an entry listing the countries the merchant delivers to. An Indexing status shows against a source while it is processed. The agent learns like a human teammate from a connected website, docs, past conversations and processes. It also reads real time customer, order and catalog data from connected systems as each conversation runs.

Sourcekonvoai.com AI platform section, Knowledge hubread 2026-09-05

The agent handles every conversation, and the team steps in only when it matters most. It resolves a conversation end to end or escalates when needed, alerting staff when it needs more information and handing conversations to the human team when it needs help. In one example a customer asks whether a product is safe with a nut allergy. The agent replies that it is looping in a teammate, and the inbox flags that the agent needs help, with full context kept across every channel and agent.

Escalate to an agent is one of the configurable general skills alongside close conversation and start a side conversation, so an operator decides when the agent hands off. Konvo names no queue before sending, draft review mode, confidence threshold or sign off for each action. Once a skill is enabled, the agent cancels orders, issues refunds and edits shipping addresses on its own.

Sourcekonvoai.com AI platform and FAQ sectionsread 2026-09-05

There is no trust center or security page, and Konvo names no SOC 2, ISO 27001 or equivalent attestation. As a German company, Konvo publishes an imprint, a privacy policy, terms and conditions and a data processing addendum, all linked from every page. The addendum sets out GDPR obligations. Conversations are assigned to agents and teams and claimed by them, but Konvo names no roles, permissions, SSO or SAML.

Sourcekonvoai.com footer legal tier and site navigationread 2026-09-05

Analyze is a named stage of the lifecycle, with a Dashboard behind it. The Dashboard uncovers opportunities, flags missing information so the agent keeps improving, and reports agent performance and customer behavior through AI driven insights. Konvo reports 2.9 million AI conversations exchanged and a 70 percent average resolution rate.

It names no way to inspect an individual run, see which Knowledge hub source grounded a reply or which skill fired, check what was written back to Shopify or the 3PL, or learn why the agent chose a refund over an exchange. Konvo also names no trace of each action, decision log or audit surface.

Sourcekonvoai.com AI platform sectionread 2026-09-05

Full context is kept across every channel and agent, so a thread that starts on Instagram and continues on WhatsApp does not restart. Recommendations draw on context, intent and past behavior, which suggests something about a shopper is retained beyond the current exchange.

Konvo does not say what the agent retains about an individual customer between separate contacts. It gives no scope per shopper, order or channel, no lifetime, and no way for an operator to inspect or clear it. Past conversations are ingested into the Knowledge hub as part of its corpus, and Konvo describes no separate memory layer.

Sourcekonvoai.com AI platform sectionread 2026-09-05

The service is hosted and reached through a web app. It is distributed heavily as a Shopify app and embedded into a merchant's existing helpdesk. Konvo names no region, region selector, VPC, single tenant arrangement, on premises build or deployment in the customer's own tenant. No hosting location is published. Konvo GmbH is a German company with a European legal tier, but that says nothing about where data sits. A data processing addendum in the footer governs how personal data is handled.

Sourcekonvoai.com homepage, pricing and footerread 2026-09-05

Prebuilt skills come grouped by domain, and a merchant switches them on instead of building them. Order management covers get order status, cancel order, edit order shipping address and refund order. Subscription management covers manage subscription deliveries and edit subscription shipping address. General handling covers close conversation, escalate to an agent and start a side conversation.

Because merchants adopt these as delivered, Konvo promises most customers are live in under a week. Recommendations arrives the same way, with Konvo learning the catalog and guiding customers to the right products without the merchant building a recommender. These are skills of one agent, not separate agents.

Sourcekonvoai.com AI platform sectionread 2026-09-05

Work arrives over six channels. They are live chat, email, WhatsApp Business, Instagram DMs, Facebook Messenger and voice. The same intelligence, personality and resolution capability runs at every touchpoint, not a thinner bot on the smaller channels. Deployment runs two ways, through Konvo's own multichannel inbox or inside an existing helpdesk.

There Konvo operates as another agent in Gorgias, Zendesk, Front or Freshdesk, so work also reaches it through the queue the merchant already runs. The agent can also start conversations itself. Broadcasts and Flows are AI powered campaigns and automations that win back past buyers through automated marketing, with unlimited no code flow automations. Konvo names no event catalog, schedule or webhook trigger.

Sourcekonvoai.com AI platform and FAQ sections, Konvo Shopify App Store listingread 2026-09-05

The five build components are Knowledge, Persona, Skills, Processes and Recommendations, and none of them sets the model. Persona covers personality and tone of voice, not the model underneath. Konvo names no provider or version, no way to bring your own key, no model assignment per skill and no routing across providers. The model looks like an implementation detail Konvo holds, since the product is sold on being onboarded in minutes with a Conversational AI Expert doing the tuning.

Sourcekonvoai.com AI platform sectionread 2026-09-05

Konvo publishes no API, SDK, developer documentation or MCP endpoint for its own platform, and the site has no developer section. Every interface runs inward, with Konvo connecting to a merchant's shop system, WMS, 3PL, CRM and helpdesk. Konvo also operates as another agent inside Gorgias, Zendesk, Front or Freshdesk while their reports, metrics and rules stay in place, so it integrates into another platform instead of exposing its own. Drag and drop flow building and platform connections do not extend Konvo itself.

Sourcekonvoai.com homepage and FAQread 2026-09-05

The platform runs in four stages, Build, Evaluate, Launch and Analyze, and Test makes up the whole Evaluate stage. A merchant simulates customer conversations to see how the agent handles real scenarios. They define typical conversations and run the full test suite automatically before deploying, then give feedback in the chat so the agent improves.

The suite tests the merchant's own agent configuration before it reaches shoppers. A Conversational AI Expert is assigned during onboarding to optimize results over time. Konvo names no pass rate, score or threshold for a suite run, and no holdout or controlled experiment after deployment.

Sourcekonvoai.com AI platform sectionread 2026-09-05

The agent reaches external systems through credentialed integrations with shop platforms, 3PLs, CRMs and helpdesks. Its skills are structured actions shaped like API calls, such as cancel order, refund order and edit shipping address, not interface operations. Konvo's own surfaces are a multichannel inbox and an embed inside the merchant's helpdesk. The agent can also call a courier or email a provider to reach a third party with no integration. It does not operate browsers, desktops or remote computers.

Sourcekonvoai.com AI platform sectionread 2026-09-05