{"slug": "ando-launches-agent-native-messaging-with-20m-to-fix-slack", "title": "Ando Launches Agent-Native Messaging With $20M to Fix Slack", "summary": "Sara Du, who built MCP servers at Anthropic in 2025, launched Ando on September 24, 2026, a team messaging platform where AI agents join channels, send DMs, and participate in live audio and video calls as workspace members, backed by $20 million from Accel, Index Ventures, Emergence Capital, and Contrary Capital. Ando is built on three layers — an Agent Execution Layer for agent identity and permissions, a Human Messaging Layer for channels, DMs, threads and calls, and a Memory Layer that tags every message with intent and context — and is agent-agnostic, letting teams run OpenAI Codex, Claude, Grokbot, or custom harnesses via a public API for searching messages, reading call transcriptions, sending messages, and managing outbound webhooks. The platform is in early access across software, real estate, and finance teams in 15 countries, positioning itself against Slack, which Salesforce acquired for $27.7B, and Microsoft Teams.", "body_md": "Slack’s AI agent model is a workaround pretending to be a feature. You @-tag a bot, wait for it to respond, lose the thread, then explain the context again. Sara Du built MCP servers at Anthropic in 2025 watching companies try to fix this with more integrations. Her conclusion: the problem is architectural, not a configuration issue. On September 24, 2026, she launched **Ando** — a team messaging platform where AI agents join channels, send DMs, and participate in calls the same way your colleagues do. Accel, Index Ventures, Emergence Capital, and Contrary Capital backed the launch with **$20 million**.\n\n## Agents as Teammates, Not Bots\n\nIn Ando, an agent is a workspace member. It has its own identity, its own inbox, and its own permissions. It can browse channels, choose which ones to join, and participate in conversations without waiting to be @-tagged. If it detects something that warrants a human’s attention, it sends the message — it does not wait for a trigger. During Jams, Ando’s live audio and video calls, agents participate in real time alongside their human counterparts.\n\nThis is not a subtle distinction. In Slack today, Claude’s broader agentic capabilities are unavailable inside the platform. ChatGPT in Slack cannot continue a shared thread with teammates or pick up context from a conversation it was not explicitly looped into. Every agent interaction requires a human to initiate it. Ando removes that requirement entirely.\n\n## The Architecture That Makes It Work\n\nAndo is built on three layers designed to solve the agent context problem:\n\n- **Agent Execution Layer** : manages agent identity, permissions, and the ability to act proactively without human prompting\n- **Human Messaging Layer** : channels, DMs, threads, and calls — a familiar surface with no new patterns to learn\n- **Memory Layer** : every message tagged with intent and context, not just content\n\nThe memory layer is the key technical bet. Feeding an agent a raw Slack export burns through context windows fast and surfaces mostly noise. Ando tags each message with what it means and why it was sent — so an agent retrieving context about a deployment decision gets the decision and the reasoning, not 300 messages of thread noise around it.\n\nAgents can also develop cross-conversation awareness: if the same problem is being discussed in three different channels simultaneously, an agent in Ando can notice that and flag it. Slack’s architecture makes this practically impossible without custom glue code.\n\n## Developer Access and Agent-Agnostic Design\n\nAndo is agent-agnostic. Teams bring whatever agents they are already running — OpenAI Codex, Claude, Grokbot, or custom harnesses. The platform does not mandate a model or a runtime. There is a [public API](https://docs.ando.so/developers/api-reference) that supports searching messages and conversations, reading call transcriptions, sending messages, and managing outbound webhooks for live event streams.\n\nThat agent-agnostic positioning matters. A platform that locks teams into its own model would die in enterprise procurement. Ando’s value is the workspace infrastructure — the identity layer, the memory layer, the context pipeline — not the model running on top of it.\n\n## The Competitive Reality\n\nSlack will respond. Salesforce acquired Slack for $27.7B partly to build AI into enterprise collaboration. Microsoft Teams has GitHub Copilot integration and Azure AI infrastructure behind it. These are not companies that cede categories quietly.\n\nThe question is not whether Slack can add proactive agents — it can, and it will. The question is whether architectural debt translates into meaningful gaps at enterprise scale. Du is betting it does, and so are Accel and Index Ventures, two firms that understand enterprise collaboration software from the inside.\n\nAndo is currently in early access across software, real estate, and finance teams in 15 countries. It is not a finished product — it is a thesis about what the next layer of workplace infrastructure should look like. For teams building seriously with agents right now, the [official launch announcement](https://www.globenewswire.com/news-release/2026/09/24/3368344/0/en/ando-launches-agent-native-messaging-platform-announces-20-million-seed) and the [TechCrunch breakdown](https://techcrunch.com/2026/09/24/ando-eyes-slack-as-it-builds-team-messaging-platform-for-humans-and-agents-to-work-together/) are worth reading. The architecture decisions being made in tools like Ando today will determine what working with AI actually means in two years.", "url": "https://wpnews.pro/news/ando-launches-agent-native-messaging-with-20m-to-fix-slack", "canonical_source": "https://byteiota.com/ando-agent-native-messaging-platform/", "published_at": "2026-09-25 18:09:49+00:00", "updated_at": "2026-09-25 19:01:27.726332+00:00", "lang": "en", "topics": ["ai-agents", "ai-products", "agent-protocols", "artificial-intelligence", "ai-startups"], "entities": ["Ando", "Sara Du", "Anthropic", "Accel", "Index Ventures", "Emergence Capital", "Contrary Capital", "Slack"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ando-launches-agent-native-messaging-with-20m-to-fix-slack", "markdown": "https://wpnews.pro/news/ando-launches-agent-native-messaging-with-20m-to-fix-slack.md", "text": "https://wpnews.pro/news/ando-launches-agent-native-messaging-with-20m-to-fix-slack.txt", "jsonld": "https://wpnews.pro/news/ando-launches-agent-native-messaging-with-20m-to-fix-slack.jsonld"}}