DDelight.ai is a branded AI concierge platform by Sendbird that lets mid-market and enterprise brands deploy a personal AI agent across customer touchpoints. It runs on Sendbird infrastructure and interacts with customers through chat, SMS, email, and voice. Its three pillars—persistent memory, omnichannel continuity, and Trust OS governance—let the AI remember past conversations, stay context-aware across channels, and operate within brand rules. Agent Steward (launched 2026) handles long-running issues, and the SaaS offering with custom pricing via AWS Marketplace aims to strengthen loyalty and engagement, turning customer experience into a competitive advantage.
Work here?Claim your company
Industries
Company Size
51-200
Company Stage
N/A
Total Funding
N/A
Headquarters
N/A
Founded
2024
See people who can refer or advise you
Help us improve and share your feedback! Did you find this helpful?
Delight.ai, Sendbird's portfolio of customer agents for enterprises, has launched Delight Customer Agent and Midas. The first is a single AI agent that handles shopping, booking, support, and retention. The second lets teams build, test, and improve that agent using natural language. The company says separate AI agents in each department split the customer journey. A Gartner survey found only 4% of customer service organisations have achieved the ROI they expected from AI initiatives. Midas lets teams make changes without engineers or IT. "Instead of waiting days for an update, those closest to the customer can make necessary changes in hours," said Nick Singh, head of product at Delight.ai. BJ's Wholesale Club, Coupang Eats, and Mixpanel use delight.ai.
Delight.ai, the customer AI agent company powered by Sendbird, has appointed Nick Singh as vice president of product. He will oversee product strategy and development, with a focus on its AI agent capabilities and helping enterprise customers grow revenue. Singh was most recently head of product at Personio, a European HR software company. He previously led growth and product development at Rippling, and spent more than five years at Google, working on Google One and advertising strategy. Delight.ai is expanding beyond customer messaging. Its AI agents aim to drive purchases and upsells, and recover stalled transactions. Over the next year, Singh will unify the company's capabilities into one platform. "Nick brings the experience and execution focus to translate that vision into products that solve real customer challenges," said John Kim, chief executive officer and co-founder.
Introducing Universal Telephony: contact center integration. September 25 2026 For many businesses, the hardest part of adopting voice AI is the phone system around the conversation. Most don't run one clean setup: a cloud contact center might handle one region and desk phones another, while a third still runs on an on-premise PBX (private branch exchange), the in-house system that routes an office's calls. Before anyone tests how an agent sounds, two questions come first: can it reach the phone system, and where does a resolved call's record go? Treating those as one question is how voice projects stall partway through, with calls connected and case history left with nowhere to land. That's why Delight is introducing Universal Telephony for Delight voice agents. Every call a Delight agent handles travels as a standard phone call, from the first ring to the handoff, and the agent can: * Answer and place calls on your own numbers, through Twilio, Telnyx, and local carrier partnerships. * Hand calls to your queues and agents by phone number, with a ticket attached where a helpdesk or contact center integration exists. * Send each resolved case to connected systems such as Salesforce, Zendesk, and ServiceNow. Answer on your numbers, hand off by phone number. Delight voice agents answer and place calls through Twilio, Telnyx, and regional carrier partnerships. Each phone number pairs with its own dedicated connection, so a business can route its existing numbers directly through the agent. When a call needs a person, the handoff is a standard phone transfer rather than a vendor-specific connection, so the agent hands callers into queues and agents by phone number. That includes Genesys Cloud and Amazon Connect contact centers, Zendesk Talk and RingCentral lines, and Zoom Phone desk devices, where calls are routed by department. Send the case where the team already works. Call records are only valuable when directed to the places your team monitors. Delight connects with platforms such as Salesforce, ServiceNow, Zendesk, Intercom, and NICE CXone, placing call outcomes alongside existing customer cases rather than locking them away in isolated silos. Think of it as a postal delivery process. The phone connection acts as the courier bringing the parcel to the building, while the integrations function as the mailroom dispatching it directly to the intended desk. Start voice from the agent you already have. For a business already running Delight as a text AI agent, adding voice is an extension of work that's already done, not a new project. Much of the setup behind the text agent carries over, so the voice agent builds on that foundation instead of starting from zero. Where Universal Telephony helps. The contact center that routes by queue. For a team running Genesys Cloud or Amazon Connect, the agent answers on the team's numbers and hands calls into its queues by phone number. Where the contact center integration is connected, the agent's summary arrives as a ticket alongside the call. The helpdesk that holds every case. A team that sets up its phone line but never connects its helpdesk can end up with calls that transfer and records that go nowhere. Connecting the helpdesk alongside the phone setup means every resolved call lands next to the customer's other cases. From the first ring to the case record. Universal Telephony gives a Delight voice agent a complete path through a business's calls. It answers on the business's own numbers, hands callers to the right queue or person as a standard phone transfer, and puts each resolved case in the connected helpdesk or CRM. The call and its record arrive together, so the team can act the moment the call ends. See how Universal Telephony works with your phone and contact center setup on the Delight.ai Voice page.
Delight.ai has launched Delight Agent Model Context Protocol (MCP), enabling AI assistants like Claude, Codex, and Cursor to access live customer service data through natural language queries. The integration allows teams to ask questions such as "Why did customer satisfaction drop this week?" and receive answers from operational data rather than dashboards or reports. The capability provides access to more than 35 operational tools, including agent performance monitoring, conversation search, quality monitoring, and write operations for updating configuration. Users can check resolution rates, CSAT scores, and search full conversation transcripts. "Enterprise software is becoming something you talk to, not something you log into," said John Kim, CEO and co-founder of Delight.ai. The feature was developed using the company's internal workflows before general release. The Delight Agent MCP is available today.
Delight.ai launches MCP: AI access to live customer data. 4h ago · 0:00 listen · Source: PR Newswire Summary. Delight.ai has launched a new integration called Delight Agent Model Context Protocol, or MCP. This allows external AI assistants like Claude and Cursor to securely access live data from a customer's Delight.ai workspace. Here's the thing: teams can now ask AI assistants questions about their customer service data using natural language. For example, they can ask "Why did customer satisfaction drop this week?" and get answers based on real-time operational data. This means moving beyond standard dashboards or exported reports. What's interesting is this also works in reverse. Earlier this year, Delight.ai allowed customers to use their own MCP servers as tools. Now, Delight.ai publishes its own MCP for external AI clients. This gives organizations direct access to over 35 operational tools, like monitoring agent performance and searching conversation transcripts. The bottom line is this aims to make enterprise software something you talk to, rather than log into, by enabling AI assistants to interact directly with applications. This is an AI-generated audio summary. Always check the original source for complete reporting.
Find jobs on Simplify and start your career today
Don't see your dream role? Check out thousands of other roles on Simplify.Browse all jobs
Industries
Company Size
51-200
Company Stage
N/A
Total Funding
N/A
Headquarters
N/A
Founded
2024
Find jobs on Simplify and start your career today
Don't see your dream role? Check out thousands of other roles on Simplify.Browse all jobs