Overhear any Silicon Valley coffeeshop conversation or onstage discussion at a tech conference these days and the (non) word “agentic” is sure to be mentioned. At BCV, we absolutely believe in a future where all apps evolve into intelligent agents. They will operate independently on your behalf, while you’re focused on other things.

But what about today? Are there any areas where we can see agents in action in our world currently?

The logical place to build agents first is in customer support. We already have virtual chatbots, but they don’t work well. Mainly, customers try to bypass them as quickly as possible to get to a human because their responses are canned and almost deliberately obtuse. They give you a menu of responses and if your issue falls outside of that narrow window, you’re out of luck until a real person can resolve your challenge. At best, it’s inefficient; at worst, hugely frustrating.

Generative AI can change that by turning bots into agents that are much more intelligent and flexible. Decagon’s agents are trained much like a human, on a company’s products, policies, and customer support history. They’re then able to automatically resolve the vast majority of customer issues, in a fraction of the time taken by manual processes. That’s what’s happening today at Decagon’s customers, which include some of the most forward-thinking names in tech like Notion, Bilt, Rippling, Substack, Duolingo, Eventbrite and Curology, among many others.

–

Like many of our partnerships, our relationship with Decagon evolved over a long period of time. I backed Ashwin’s prior company, Helia (acquired by Scale), and through him got to know Jesse. We quickly found they’re kindred spirits: two, second-time founders with that rare combination of intellect, intensity, and customer-obsession. Suggest to Jesse that something will take two weeks and he’ll push to get it done in three days.

As we’ve been informally advising Decagon over the past year or so, we’ve been struck by a several notable things:

1. Working with Decagon meaningfully impacts companies, who have been able to scale their teams by the equivalent of dozens of CS agents, saving millions of dollars each year.
2. Companies love the transparency Decagon offers, as well as the near-constant fine-tuning. Decagon’s product improves its answers in real-time, solving problems without help, but it also provides insight into “how” it arrived at its answers so that companies can update their documentation as needed.
3. Companies (and their customers) love this product! Nearly every time Decagon has gone head-to-head with competitors, they’ve landed the deal. This is because of Decagon’s unique approach which offers greater accuracy and transparency into the “why” behind the answers it gives.

–

Decagon agents are working today in chat, but what’s next? Voice is a logical next step and the company is working on that new modality. It will require even greater precision and the ability to carry on an even more “human” conversation, imbued with empathy and care.

We see customer support as just the starting point for this technology. Over time, agents will expand from helping existing customers to every prospect that engages with a company. Imagine that each customer has an agent that acts as a personalized concierge, bringing them into the storefront and showing them products tailored to their past history. This superpowered agent would know them by name at first contact, make helpful suggestions and proactively solve their problems. Every store would be able to offer Four Seasons-level service that would engender loyalty on a new scale.

We’re incredibly excited to be a part of building this future with Jesse, Ashwin and the entire Decagon team.