Lorikeet is an AI-powered customer support agent built for companies in complex, regulated industries such as fintech, healthtech, and energy. Unlike simple chatbot deflection tools, Lorikeet is designed to take real actions inside connected systems — processing refunds in Stripe, checking order status via internal APIs, or updating records in Zendesk — following the same standard operating procedures (SOPs) that experienced human agents use. The platform uses a dual-agent architecture: a Concierge agent that handles customer tickets end-to-end across channels (chat, email, SMS, WhatsApp, and voice), and a Coach agent that helps support teams configure, test, and continuously improve workflows. Rather than relying on a single large language model to generate responses, Lorikeet uses an orchestration layer that marshals LLMs for specific subtasks while a deterministic graph engine tracks conversation state and enforces business logic. Lorikeet integrates natively with help desk and CRM platforms including Zendesk, Intercom, Front, HubSpot, Shopify, and Stripe, as well as knowledge bases such as Confluence, Notion, Google Workspace, and Guru. It also supports custom tool connections via APIs, webhooks, and MCP servers, enabling it to reach proprietary internal systems. Escalation to human agents is built in for cases the AI cannot confidently resolve. Founded in Australia and now headquartered in San Francisco, Lorikeet raised a $5M seed round in 2024 and a $35M Series A in August 2025 led by QED Investors, with participation from Blackbird, Square Peg, Airtree, and angels including founders of Canva and Airwallex. The company reached a valuation above $200M on over $54M in total funding, with customers including Airwallex, Linktree, Flex, Eucalyptus, and Magic Eden. Key features: - Dual-agent architecture: Concierge (resolves tickets) and Coach (configures and improves workflows) - Executes real actions in integrated systems such as processing refunds in Stripe or updating Zendesk records - Follows human-authored standard operating procedures (SOPs) rather than purely generative responses - Omnichannel support across chat, email, SMS, WhatsApp, and voice - Integrations with Zendesk, Intercom, Front, HubSpot, Shopify, Stripe, Twilio, Slack, and Sunshine Conversations - Knowledge base connectors for Confluence, Notion, Google Workspace, and Guru - Custom tool connections via APIs, webhooks, and MCP servers - Seamless escalation to human agents for unresolvable or sensitive cases
Credit-based, per-resolution pricing. Simple FAQ answers cost 0.67 credits; complex workflow resolutions cost 1.25 credits. Exact plan tiers and credit bundle prices are not publicly listed.