About
Duckie builds an agentic AI platform that lets customer-support teams create agents to resolve tickets, investigate issues, triage escalations, and automate workflows. It sells primarily to support organizations, especially B2B SaaS companies, and differentiates through plain-English, no-engineer workflow creation combined with production safeguards such as testing, policy enforcement, human approvals, audit traces, and kill switches.
Market
Duckie competes in generative-AI customer support and customer-service operations software, positioning itself as an autonomous ticket-resolution platform for AI-first and B2B SaaS support teams. Its differentiation is that it performs investigative and transactional actions—such as refunds, account updates, password resets, and subscription changes—and can build agents from a company’s existing knowledge and ticket history, rather than functioning only as a FAQ or ticket-deflection layer. It also connects to an existing helpdesk and broader developer-tool stack instead of requiring customers to replace their support system.
Duckie primarily targets B2B SaaS and other AI-first product companies with customer-support teams handling high-volume or technically complex tickets. Its buyers are support and operations leaders seeking to automate resolution and reduce hiring or engineering dependence, rather than merely deflecting questions.
At a Glance
Problem
Customer-support teams face a costly combination of rising ticket volume, repetitive inquiries, fragmented internal knowledge, and pressure to respond instantly without expanding headcount. Duckie’s documentation frames the economic payoff as handling up to 10 times the volume with the same team, reducing response times from hours to seconds, and freeing agents for complex, high-value conversations. Its clearest use case is end-to-end resolution of operational support requests—such as plan changes, refunds, account updates, and technical investigations—rather than merely deflecting simple questions.
Product / Service
Duckie is an agentic customer-support platform that lets support teams build and operate AI agents without engineers. Teams connect channels such as Zendesk, Intercom, and Slack; connect knowledge sources such as Notion, Confluence, and Google Drive; and describe workflows, tone, rules, and safety guardrails in plain language. The agents search the knowledge base, follow runbooks, take actions through integrations such as Jira, Zendesk, Slack, Stripe, and other tools, respond in the originating channel, and escalate cases to humans when appropriate.
The benefit is a production support layer that can resolve tickets and execute workflows autonomously while retaining oversight through testing, traceability, permissions, approval rules, and a kill switch. This combines faster and more consistent customer responses with lower support workload and the ability for human agents to spend more time on difficult cases.
Market
Duckie competes in the fast-growing market for AI customer-service agents and agentic support automation. Comparable platforms in the category include Fin, Ada, Decagon, Sierra, Kore.ai, Cognigy, Zowie, and Gorgias, while incumbent support systems such as Zendesk and Intercom are also relevant channel and workflow ecosystems. Duckie differentiates itself by positioning support teams as builders of multi-step agents that resolve tickets, investigate issues, and perform backend actions, rather than limiting AI to answer generation.
The company has evidence of early commercial traction: its site names customers including Automox, Grid, Vanquish Trader, Abbiamo, and Medtelligent; it reports an Automox team handling 30% more tickets with one support person; and a published case-study listing says Grid resolves 79% of tickets with Duckie. YC identifies Duckie as founded in 2023 by Valerie Li and Joel Ritossa and having four San Francisco employees, indicating an early-stage company. The available evidence does not disclose revenue, so Duckie should not be definitively labeled pre-revenue; the strongest supported conclusion is that it is an early-stage, launched and customer-backed company with no public revenue figure in the reviewed materials.
Founders & Leadership
Funding History
Y Combinator
Recent News
A startup profile describes Duckie as an AI support-agent company serving B2B SaaS and technical products. It also notes early traction from integrations with Slack, Jira, and Confluence.
Duckie published a case study reporting that it handles 95% of Vanquish Trader’s support tickets with high accuracy, substantially changing how the company’s founders manage support.
Duckie announced its partnership with options prop-trading firm Vanquish Trader. The company said the integration automated 95% of tickets and reduced support costs by more than 50%.
Duckie announced that it would sponsor HumanX, staff a booth, and host a private happy hour in San Francisco for attendees and prospects.
Duckie announced its participation in Support Driven Summit LA and said it would be present throughout the event while hosting a dinner for support and CX leaders.
Duckie’s updated documentation describes the platform as an AI-powered customer-support product that automates conversations, resolves tickets, and helps support teams work more efficiently.
Duckie reported that Grid fully resolves 79% of incoming tickets without agent involvement, with 75% of those tickets completed in under five minutes across more than 15,000 monthly tickets.
Duckie promoted a no-code approach for building support agents, claiming deployment in two weeks with 90%+ resolution rates. The article highlights end-to-end workflows such as refunds, password resets, and ticket resolution.
Duckie argued that companies can deploy self-building AI support in roughly two weeks rather than undertaking a six-month setup, while automating workflows such as refunds, password resets, and ticket resolution.
Latka’s company profile described Duckie as an AI assistant for customer-support teams and reported estimated 2025 revenue of $440,000.
Active Roles
2Business Model
Duckie appears to monetize through paid SaaS subscription plans. A 2026 pricing review reports four plans starting at $725 per month when billed annually, with limits scaling by messages, workflows, agent tools, and knowledge searches.