Trace: AI Workflow Automation for Human-AI Teams

This podcast announces the launch of Trace, a workflow automation platform designed to integrate human and AI agents. Trace aims to streamline routine tasks by connecting various tools like Slack, Jira, and Notion, identifying opportunities for automation, and assigning repetitive work to AI agents. The platform creates a unified company index by analyzing existing data (wikis, documents, chats) to provide rich context for task routing and workflow generation. Users can either leverage existing templates or create custom AI agents, with the system learning from user preferences to optimize the human-AI task division, while also offering a dashboard to track active workflows.

The big picture is made up of millions of tiny, focused actions; be faithful in the small things

Guids Hub

Frequently Asked Questions

What is Trace and what problem does it solve?

Trace is a workflow automation platform designed for a “human 👾 AI workforce.” It routes tasks to the appropriate agent, whether human or AI, by connecting with existing tools like Slack, Jira, and Notion. The core problem it addresses is the time and focus drained by routine, repetitive work such as chasing updates, filling documents, and coordinating teams. By identifying automation opportunities and deploying AI agents for these repetitive tasks, Trace aims to streamline operations and free up human teams for more complex work.

How does Trace integrate with existing company systems and leverage data?

Trace integrates with a company’s existing systems to analyze various data sources, including wikis, documents, chats, activity logs, and people data. This comprehensive analysis builds a “unified index” of the company’s work, providing rich context for every new workflow. This rich context is crucial for effectively routing tasks to the right human or AI agent, ensuring that AI agents don’t operate in silos but rather with a full understanding of the company’s operations and information. This unified context layer is seen as a “game-changer” for making AI agents truly effective.

What are the key functionalities and features offered by Trace?

Trace provides several key functionalities to automate and manage workflows:

• Workflow Generation: Users can generate workflows from a single prompt and track the entire project lifecycle.
• Task Breakdown and Assignment: It breaks down tasks into steps and assigns them to the most suitable agents, human or AI.
• AI Agent Deployment: Users can deploy AI agents for various tasks using pre-built templates or by creating their own custom agents.
• Automated Workflow Execution: Workflows can run automatically with real-time triggers and schedules.
• Dashboard for Visibility: Trace offers a dashboard to track all active workflows, with future enhancements planned to include AI Agent performance, task state trackers with alarms, debug capabilities, and anomaly tracking.

How does Trace determine whether a task should be handled by an AI agent or a human?

Trace employs a set of internal rules to decide whether a task should be routed to an AI agent or a human. These rules consider factors such as the type of workflow, the number and complexity of tasks, and other parameters. While the system currently makes these decisions, there’s interest in potentially allowing teams to customize this logic. Furthermore, Trace learns from user behavior; if a user consistently opts to handle a specific type of task themselves rather than allowing AI to automate it, the system will adapt and respect that preference, putting users in control.

Can Trace handle complex, multi-platform content distribution workflows?

Yes, the sources suggest that Trace is capable of handling complex, multi-platform content distribution workflows, even those requiring significant customization. An example provided is automating the process of taking a video, generating a YouTube-optimized title and description, and then adapting it for various platforms like TikTok, Instagram, and Facebook, including title rewrites, description adjustments, and hashtag generation. While such a workflow might initially require some “prompt engineering” for each task, the platform’s makers acknowledge this as a strong use case and are exploring features like direct video upload to further simplify such processes.

What sets Trace apart from other automation tools, especially those that “break”?

The makers of Trace specifically highlight its ability to keep things running reliably, unlike other attempts to “stitch together” tools like Jira, Notion, and Slack using general automation platforms (like “zaps”) which often break. The key differentiator is Trace’s unified context layer, which gathers and analyzes data from all connected systems. This rich, integrated context prevents AI from feeling “broken” or working in silos, as it provides a comprehensive understanding that generic automation tools might lack. This holistic view is what makes AI agents click and function effectively.

Is Trace accessible to non-technical users, and what is its business model?

Yes, Trace is designed with a user interface that a non-technical person can understand, making it accessible to a broader audience within an organization. Regarding its business model, Trace offers “Free Options,” indicating that there is a free tier or trial available to users. This allows potential customers to sign up and experiment with automating their workflows.

What are the future development plans for Trace, particularly concerning its dashboard and AI capabilities?

Trace has ambitious plans for future development. While a basic dashboard for tracking active workflows currently exists, the team is working on making it much more comprehensive. Upcoming features for the dashboard include:

• Access to AI Agent performance metrics.
• Advanced task state trackers with alarms and debugging capabilities.
• Task anomaly trackers to monitor issues like “stale” tasks. The goal is to evolve towards an “agentic infrastructure that learns from your company and introduces automation without human intervention,” showcasing a long-term vision for increasingly intelligent and autonomous workflow management.

Check The Product