OpenTag

Teams run work across many disconnected tools like Slack, Notion, Linear, GitHub, Drive, CRM, and analytics platforms, and keeping track of decisions, answering recurring questions, and completing routine tasks requires manually searching across all of these sources. When teammates leave, the knowledge they held in conversations and documents often leaves with them, forcing new hires to ask around rather than find documented answers. Delegating work to a teammate normally means sharing account access or context, and any model-based assistant tied to a single AI vendor is both locked into one model's limitations and expensive if every task is sent to a frontier model regardless of complexity.

OpenTag is an AI coworker that users tag directly in Slack threads, channels, or DMs to get answers or trigger actions, with Microsoft Teams support coming soon. It connects to company systems including Google Workspace, Notion, Linear, GitHub, PostHog, Stripe, Reevo CRM, Granola, Circleback, Cal.com, Figma, Meta Ads, LangSmith, and over 32 more tools, and runs secure cloud agents scoped only to what a task needs, ending each run when work completes. Actions with real consequences are surfaced for approval before changes happen, recurring requests can be converted into scheduled or event-triggered automations, and permissions are limited to whatever the requesting teammate already has access to, so private channels and data stay private. OpenTag also builds and updates a company wiki from conversations and connected sources, tracking changes with sourced citations, and routes each request across more than 40 models from providers like OpenAI, Anthropic, Google, and xAI to the cheapest model capable of the task, targeting up to a 70% cost reduction versus always using a frontier model.