Andrew Bolis on Building AI Teammates for Workflow Automation

In a recent LinkedIn post, Andrew Bolis explores how businesses can leverage Artificial Intelligence (AI) to build agents that function as true teammates, capable of managing complete workflows rather than just responding to prompts.

Andrew Bolis highlights a significant gap in current AI capabilities, noting that while many tools can generate content or answer questions, they fall short in managing the complexities of ongoing workstreams. He points out the manual effort still required for tasks such as updating project trackers, creating follow-up actions for team members, and notifying stakeholders about blockers. “Teams waste hours on handoffs and constant supervision,” Andrew Bolis states, emphasizing the inefficiency inherent in these manual processes.

Bridging the Gap with AI Teammates

To address these challenges, Andrew Bolis introduces the concept of “ClickUp Super Agents.” These AI agents are designed to understand project goals, develop execution plans, and coordinate across various tools and data sources. A key feature highlighted is their ability to maintain context from previous work and continuously improve their performance over time.

“They’re AI teammates that understand goals, create execution plans, and coordinate across tools and data.”

Andrew Bolis elaborates on the cross-departmental applications of these AI teammates. In sales, Super Agents can route leads and update CRM records. For DevOps, they can monitor deployments and update release documentation. Marketing teams can benefit from agents that track campaign performance and flag issues, while Customer Success agents can detect churn signals and create outreach tasks.

A Step-by-Step Guide to Building AI Agents

The post provides a practical, step-by-step guide for building these AI agents within the ClickUp platform, emphasizing that no specialized tech skills are required. Andrew Bolis outlines the process as follows:

1. Set your agent’s goal: Users describe the desired outcome in plain language, such as “Send daily blocker summaries.”
2. Connect your tools and data: The agent is given access to relevant workspace areas, tasks, documents, and integrations.
3. Define actions and workflows: Specific actions like creating tasks or sending notifications are configured, along with triggers based on schedules or task changes.
4. Set permissions and approval rules: Controls are established for where the agent can operate, with an option for an “Approval Mode” for sensitive actions.
5. Test and launch your agent: The agent’s performance is validated through testing before being launched and continuously monitored for refinement.

Once live, Andrew Bolis suggests treating these agents like human teammates, assigning them tasks directly or mentioning them in conversations. The agents then execute the work and notify when it’s complete.

“Build AI teammates that plan, execute, and improve workflows automatically.”

Andrew Bolis concludes by encouraging teams to embrace this technology to automate workflows and free up human capital for more strategic decision-making. As he puts it:

“Each agent works in the background while teams focus on strategic decisions.”

The core message from Andrew Bolis’s post is that AI agents, when properly configured as “Super Agents,” can significantly enhance team productivity by taking over routine workflow management, thereby allowing human team members to concentrate on higher-level strategic tasks.