Funding
Funding not disclosed
Founders
Product
Problem
Product development teams often run AI agents in isolation, leading to fragmented context, hidden assumptions, and late human review that can cause duplicated effort and missed errors. Manual handoffs and copy‑paste workflows make it difficult to keep agent actions, decisions, and artifacts visible to the whole team.
Solution
Vokal offers a shared workspace where human team members and AI agents collaborate in real time. Each agent is given a workspace identity with defined roles, permissions, and runtime, allowing it to operate alongside humans on a common channel. As agents execute tasks, their prompts, tool calls, intermediate outputs, and approvals stream live into the workspace, creating an inspectable run history. Humans can monitor progress, intervene for judgment or redirection, and approve results before they are finalized. Completed outcomes, decisions, and source artifacts are saved as reusable knowledge, enabling repeatable and efficient workflows across engineering, product, growth, and other functions.
Target Audience
Primary customers are product development teams—including engineering, product, growth, marketing, and customer operations—that integrate AI agents into their daily workflows and need a unified, visible collaboration environment.
Features
- Role‑based agent profiles with customizable runtime, memory, tools, and permissions
- Live status stream that displays agent reasoning checkpoints, tool calls, partial outputs, and approvals within the shared channel
- Human‑in‑the‑loop controls for real‑time intervention, approval, and redirection of agent work
- Persistent workspace record that captures prompts, files, decisions, and corrections as reusable knowledge or memory
- Support for local agents, hosted Hermes agents, MCP‑based agents, and custom ACP‑compatible stacks (e.g., Claude Code, Codex, OpenCode)
- Integrated agent capacity limits tied to team size plans, with desktop access for organization‑wide visibility