Duckie (YC W24)

Thinking about having a billboard for Duckie (YC W24). 1 or 2?

Duckie (YC W24) use case demo #2 One of the most common support tickets in SaaS: “How can I change my plan?” We built a sample workflow in Duckie to handle the entire thing automatically. It asks the user for the right info, pulls their current plan, makes the change via API, and escalates to the team when needed. If your team already has a runbook for handling some tickets, you can turn it into a workflow in Duckie. Book a demo with us to see how it could work for your team! https://lnkd.in/eJA9qTmx #customersupport #aiagent

Introducing Duckie (YC W24) no-code AI support agent builder. Most support processes already follow a clear runbook - checking systems, searching docs, taking action, escalating when needed. Duckie lets you turn that into a workflow with actionable steps that an AI agent can take to resolve tickets. In this video, I walk through how I built a Stripe refund workflow. Unlike a chatbot that just points to an article, Duckie’s agent can check your billing system and process the refund for the customer, following your runbook. Details: https://lnkd.in/eduwrtKw If you are interested, let's chat: https://lnkd.in/eJA9qTmx #ai #aicustomersupport #customersupport

Trust issues aren’t just for exes. AI agents have them too. Remember when Air Canada’s chatbot made up a fake bereavement discount, and the airline had to honor it? That’s what happens when your AI gets too creative. That’s why we built Duckie (YC W24) to be boring in the best way: predictable and trustworthy. Here's how ->

And God (once again) said - let there be SaaS! We’ll be attending #SaaStr this week in San Mateo. If you want to grab a coffee (and a cute rubber duckie swag) during the conference, grab some time here: https://lnkd.in/ektqtSHm Topics we are good at: 1. Navigate bringing in an AI support automation tool for the first time - what use cases to start with 2. How to measure the ROI of AI support tools and know if the money is well spent 3. How to pick the right support automation tool for different verticals (fintech, e-commerce, B2B, B2C) 4. Pitfalls that cause AI solutions to fail — and how to avoid them 5. Tradeoffs of building vs. buying AI tools 6. Come up with cute SaaS brand names (like "Duckie")

We're hiring our Founding GTM at Duckie (YC W24)! If you’re a hands-on GTM operator who loves moving fast and closing deals, and thrives in early-stage chaos and wants true ownership, this role is for you. We’re looking for someone hands-on, creative, with an authentic passion for sales. You will - Build outbound campaigns - Run inbound engines - Create content, host events, and explore creative channels to drive growth - Create repeatable GTM playbooks as we grow Apply here:

We are excited to announce our new hire for Duckie (YC W24) - Jinx! Title: Junior Software Engineer (she's not very good) Responsibilities: 1. Walk staff members 2x a day. 2. Provide morale boosts through unsolicited physical affection. 3. Conduct daily patrols of the office floor. 4. Test durability of office furniture through bite-driven quality assurance. If you have dog parks recs around soma, please let us know. We are planning a welcome offsite for her.

Introducing: No-Code AI Agent Builder for Customer Support Most AI support tools are one-size-fits-all black boxes — hard to customize, impossible to control, and can’t handle your team’s unique workflows. So we flipped it. Duckie gives support teams the power to build their own AI agents without code. Use our building blocks — workflow builder, responders, knowledge search, and plug-and-play APIs into tools like Stripe, Salesforce, and logs — to automate complex tickets like refunds, log investigation, and subscription issues. You know your support best, so we give you the power to take AI fully in your control. Start resolving the tickets no black-box agent can. Read more here: https://lnkd.in/euZz7xfN

Introducing: No-Code AI Agent Builder for Customer Support Most AI support tools are one-size-fits-all black boxes — hard to customize, impossible to control, and can’t handle your team’s unique workflows. So we flipped it. Duckie gives support teams the power to build their own AI agents without code. Use our building blocks — workflow builder, responders, knowledge search, and plug-and-play APIs into tools like Stripe, Salesforce, and logs — to automate complex tickets like refunds, log investigation, and subscription issues. You know your support best, so we give you the power to take AI fully in your control. Start resolving the tickets no black-box agent can. Read more here: https://lnkd.in/euZz7xfN

I get the temptation. You're smart. You’ve got great engineers. You read a few LangChain tutorials and now you’re 3 prompts away from “revolutionizing customer support.” So... should you build your own AI agent? Maybe. But it’s worth knowing what you’re signing up for. I wrote this after seeing a bunch of teams try both: - What building actually involves - What buying gets you (and what it doesn’t) - Why a hybrid model might be the most practical path Also, got some unique use cases and trying to figure out what to do? I’m down to chat and offer some unsolicited opinions. (Or just sit back and enjoy the chaos with you - both are fun.) Book on my calendar: https://lnkd.in/e38gCQUA