Decagon

What do you get when you mix a product builder, an AI whisperer, and a trusted customer advisor? ✨ You get an Agent Product Manager at Decagon. Decagon’s Agent PMs sit at the bleeding edge of how real businesses deploy AI agents to transform customer experience— from automating complex HR workflows at Rippling, to simplifying processes for Substack readers and creators, to helping ClassPass members get refunds for unexpected circumstances. From scoping use cases with CX leaders, to helping brands launch high-performing AI agents into the wild, Agent PMs bring AI to life in the real world. The full breakdown is live on our most recent blog from Max Lowenthal, Decagon’s first Agent PM. Check it out to see why this role is one of the most exciting jobs in tech right now. 🔥

Honored to be included on the 2025 Forbes AI 50 — Forbes’ annual list of the most promising privately-held AI companies in the world. At Decagon, we’ve taken a radically different approach to building conversational AI agents for customer experience. We call it Agent Operating Procedures, or AOPs. AOPs combine the power and flexibility of natural language with the precision and rigor of code, enabling businesses to build and scale AI agents that are smarter, faster, and reflect the true voice of each brand. Read more about our approach to AI CX agents in the comments 👇 #ForbesAI50

Decagon has pioneered a novel approach to building AI agents for customer experience. We call it Agent Operating Procedures, or AOPs. Today, I'm excited to share more about our vision behind AOPs and what makes the approach ideal for enterprises in 2025. The truth is: most customer service automation — including many of today's GenAI solutions — are just fancy chatbots that break in real-world situations when workflows get too complex. This isn't scalable, especially for companies handling millions of customer interactions. So why have businesses struggled to deliver great customer experience in the past, and why have these issues persisted even with some of the newer GenAI solutions? ❌ These systems are costly and slow to build, requiring technical resources and resulting in rigid frameworks that can't adapt quickly to real-world situations. ❌ They become black boxes, making it impossible to track decision-making or iterate efficiently on the AI's logic. ❌ They produce poor-quality results because when you can’t iterate quickly, you can’t reliably deliver results that are at or above what a human agent would deliver. Decagon's AOPs combine the power and flexibility of natural language with the precision and rigor of code. This innovative combination allows Decagon to deliver AI agents that are not only smarter and faster, but also more secure, reliable, and scalable. I explain more in the video below. If you have any other questions about how AOPs work, leave a comment or send me a message. Would love to show you around the product!

At Decagon, we’ve taken a radically different approach to building conversational AI agents for customer experience. We call it Agent Operating Procedures— or AOPs. ✨ AOPs combine the power and flexibility of natural language with the precision and rigor of code. This innovative combination allows Decagon to deliver AI agents that are not only smarter and faster, but also more secure, reliable, and scalable. Check out our latest video where Decagon Co-founder & CEO Jesse Zhang explains our approach, and learn more in his blog post in the comments. 👇

Welcome to our new teammates across People, Talent, and Operations! • Gabriella Parisi, People Operations • Kaly Suarez, Talent • Kyra D'Onofrio, Deployment Strategist We're so excited for these wonderful additions to the team! Speaking of scaling, we're hiring for all departments! Check out our careers page below:

A recent experience with an AI bot: I filed a support ticket with Rippling regarding a benefits question and recieved an immediate response that articulated the problem back to me clearly, provided a 24 hour SLA, and then signed the ticket "Rippling Support AI". 3 hours later a human support agent resolved and closed the ticket for me. Problem solved. This AI -> human handoff was perfect for my use case which required 1) a timely response, and 2) the option to go deeper depending on the answer. Fun to see AI use cases that are making the users life better AND making the employees life better (in this case the support rep)

We're thrilled to welcome our newest GTM members to the Decagon team! • Breuer Bass, Sales Engineer • Kyle Egan, Account Executive • Hayden Conrad, Account Executive These talented hires will be key players in supporting Decagon's expansion through 2025 and beyond. Interested in joining us? Head on over to our careers page! We're hiring across all departments. ➡️ https://lnkd.in/gwvw-VGJ

Our engineering team is growing quickly! We're so excited to welcome this talented group aboard, they're already making an impact. Shoutout to Sahib Singh Pandori and Marcus Schorow, two new senior engineers joining our technical staff! If you're looking for our next play, check out our careers page. We're hiring across all departments! ➡️ https://lnkd.in/gwvw-VGJ

One interesting topic these days is how you continuously improve AI agents once they are live in production. 🌀 Here's one way we approach it at Decagon. As our AI agent has more and more conversations with customers, it can identify areas that get escalated to a human agent or topics where the knowledge/processes are lacking. From there, the next step is to suggest improvements. You can use LLMs to read every conversation, categorize the biggest themes, and then read how the human agents solve these issues to draft new knowledge articles or new procedures for the AI agent to follow. Over time, this feedback loop will continuously improve the AI's ability to answer new questions and handle various situations. LLMs can be used both to power the AI agent itself and to drive improvements. Of course, this example is specific to customer support, but you can easily imagine how similar mechanisms can exist for other AI agents. I predict this will be an increasingly important theme as more AI agents get deployed to production.

We're thrilled to be listed as #2 on the 2025 Enterprise Tech 30 mid-stage list! 🎉 The Enterprise Tech 30 names the most promising private companies in enterprise technology with the most potential to meaningfully shift how tech enterprises operate. A huge thank you for the recognition to Wing Venture Capital, Newcomer, and all the institutional investors who participated. Check out Decagon's profile and the full #ET30 list in the comments. 👇 Excited about building the future of conversational AI agents for customer experience? We’re hiring across the board! Check out our open positions in the comments. 🚀