AI-powered enterprise communications company Sendbird this week launched a white-labeled personal concierge. Known as Delight.ai, it gives brands and app developers a way to integrate VIP-level concierge functions into their customer-facing apps, thus boosting their quality and stickiness.
Sendbird provided us with the example of ordering dinner on DoorDash. By interfacing with the concierge, powered by Delight.ai, end-users can tell the concierge in a natural conversational style that they don’t like spicy food, or have a nut allergy. Delight.ai will then remember that preference next time.
According to Sendbird, the next time you open the same app, it will greet you with a friendly message and a reminder that it has removed spicy options from the menu based on the last order. The interface itself is purposefully conversational and accommodating to evoke personal concierge vibes.
“No longer will customers have to restate preferences or experience disjointed conversations across multiple channels. It’s a true step change in how brands engage,” Sendbird cofounder and CEO John Kim told Localogy Insider. “This shifts customer engagement from a cost center to a competitive advantage, ensuring every brand can deliver experiences that make each customer feel seen, remembered, and delighted, driving lasting loyalty and lifetime value.”
Long-Term Memory
Besides its persona, delight.ai is differentiated by its memory. According to Sendbird, most AI agents are trained on static knowledge bases, while Delight.ai has a long-term memory. The intended UX is the same VIP feeling one gets when a high-end hotel or casino remembers their perks & preferences.
The other point of differentiation is scale. Because Delight.ai will operate across apps, it will have a wider corpus of knowledge on which to train itself. In the AI land grab, the name of the game – or at least one of them – is the quality and quantity of training data. Sendbird’s operational scale offers an edge there.
Meanwhile, the same operational scale could work on other levels. For example, for users who interface with several apps that have integrated Delight.ai, it could be effective in remembering preferences across apps. This is our speculation, but it seems like a natural direction for Delight.ai as it grows.
Similarly, Senbird connects more than 300 million people globally across its software. And the company’s flagship TustOS platform hovers over everything to ensure that agent behavior is compliant with things like brand-defined parameters and brand safety. This is key to resolving AI’s trust issues.
The Time is Right
The timing could also be right for Delight.ai. Despite some reticence and trust issues surrounding AI, consumers have spoken. Sendbird reports that 62 percent of consumers prefer chatbots over waiting for human agents. Meanwhile, retail chatbots are projected to grow to $72 billion in revenue by 2028.
These figures are broad but should also apply to the SMB market, albeit with slower adoption, as it often goes. With AI, SMBs will be held back by standard friction and latency. However, AI’s value proposition will break through as it aligns with common SMB sentiments like saving time and offloading rote work.
If all the above pans out, Delight.ai will be well-positioned to democratize agentic AI by easing brands and SMBs into it. Tools like Delight.ai, or any AI-as-a-service, could be an opportune subsector as demand ratchets up for plug & play tools for consumer-facing brands to offer agentic AI to their users.
“Delight.ai’s value lies in democratizing advanced customer experience,” said Kim. “By introducing the first branded, autonomous AI concierge built on persistent memory, we are fundamentally leveling the playing field.”
