Build Any n8n Workflow in Minutes, Here’s How

Key Concepts

• NA10 Workflow Builder: A new feature in NA10 that uses AI to build and automate workflows.
• AI Agents: Automated programs designed to perform specific tasks, now significantly easier to create.
• Build with AI: The core functionality in NA10 that allows users to prompt an AI to construct workflows.
• Nodes: Individual components or steps within an NA10 workflow (e.g., Daily Trigger, Set, Google Calendar, Code, AI Agent, Gmail).
• Node Parameters: Configurable settings for each node, which the NA10 AI can now pre-fill.
• Credentials: Authentication details for connecting services (e.g., Google account for Gmail/Calendar).
• Ask Mode (NA10 AI): An AI mode for understanding, explaining, and suggesting changes without directly modifying the workflow.
• Build Mode (NA10 AI): An AI mode that actively makes changes and builds/modifies the workflow based on prompts.
• 1,000 Character Limit: A current restriction on the prompt length in NA10 AI's build mode.
• Vectal: An external AI chatbot used as a workaround for longer prompts, offering access to advanced models like Anthropic Sonnet 4.5.
• Hostinger (VPS): A recommended hosting solution for NA10 agents, providing unlimited workflows and concurrent executions.
• Debugging Hell: A term describing the difficulty of fixing numerous errors in complex, multi-node workflows.
• Human-in-the-Loop: A collaborative approach where AI agents perform automated tasks, but humans provide oversight, vision, and handle tasks beyond AI capabilities.
• Temperature (AI Model Parameter): Controls the randomness of an AI model's output (0 for deterministic, 1 for random).
• Max Tokens (AI Model Parameter): Limits the length of an AI model's generated response.

---

Introduction to NA10's Workflow Builder and AI Agent Creation

The video introduces NA10's new workflow builder, which has made building AI agents "10 times easier," reducing the creation time to as little as "2 minutes." David Andre, with three years of experience building hundreds of AI agents, highlights this update as the "biggest NA10 update of all time." The new "Build with AI" button or "NA10 AI" sidebar allows users to leverage AI to construct automations from simple prompts, marking a "completely new era of building AI agents." This functionality enables the AI to search available NA10 documentation, add necessary nodes, connect them properly, and even update node parameters with pre-filled credentials. The speaker emphasizes that if businesses are not building AI agents now, they are "falling behind" due to the unprecedented ease of entry and the "enormous" opportunity, given that "tens of millions of businesses" still don't use AI.

---

Demonstration 1: Simple Email Draft Agent

The first demonstration involves building a simple agent to create an email draft.

1. Initial Build: From a single prompt, the NA10 AI built a complete agent in "30-40 seconds with zero errors." The AI performed four key steps:
   • Searched available NA10 documentation to understand nodes and connectors.
   • Added necessary nodes (initially scrambled).
   • Ordered and connected nodes properly.
   • Updated node parameters, including pre-filling credentials (e.g., Gmail account, model selection, system prompt).
2. Execution and Debugging:
   • Upon initial execution, the workflow ran successfully but produced an empty email draft, indicating a content issue.
   • The speaker manually adjusted a variable in the "Gmail" node to correctly pass the email content.
   • A subsequent error in the Gmail tool was fixed by asking the NA10 AI directly, which identified the issue as a placeholder sender email address. The speaker then manually dragged in the correct sender.
   • After these fixes, the agent successfully created a detailed email draft, a task that previously would have taken "10-15 minutes" for an experienced user or "30 minutes" for a beginner.

---

Hosting AI Agents with Hostinger

The video addresses the common question of where to host AI agents, recommending Hostinger.

• Benefits: Hostinger provides "unlimited workflows and unlimited concurrent executions" by giving users their own virtual private server (VPS). The speaker and his team use the KVN2 plan internally.
• Black Friday Deal: A significant Black Friday deal makes a VPS available for around "$65," offering "64% savings" on a 24-month plan, plus a "free domain for a year." An additional "10% off" is available using the code "David."
• Setup Process:
  1. Visit Hostinger via the provided link.
  2. Select a VPS plan (KVN2 recommended).
  3. Choose a 24-month period for the best deal and long-term agent operation.
  4. Apply the coupon code "David" for an extra 10% discount.
  5. Select the server location closest to the user (e.g., France).
  6. Select "NA10" as the operating system.
  7. Complete checkout and account creation.
  8. Manage the VPS in the Hostinger panel, where CPU usage, memory, and traffic can be monitored.
  9. Access the NA10 server via "manage app" to manage AI agents.

---

Demonstration 2: Daily Calendar Analysis and Report Agent

This section focuses on building a more sophisticated agent that analyzes daily calendar events and generates a concise report.

• NA10 AI Modes: The video introduces two modes within NA10 AI:
  • Ask Mode: Similar to chatbots like Cursor or Claude, it explains and suggests but "doesn't make changes" to the automation.
  • Build Mode: The powerful "agent mode" that "can make changes" to the workflow.
• Overcoming the 1,000 Character Limit: A significant limitation of the build mode is its "1,000 character limit" for prompts. To circumvent this, the speaker uses Vectal to:
  • Split the complex build idea into smaller chunks.
  • Build the AI agent "one node at a time."
  • Leverage Vectal's access to "latest and greatest models" (e.g., Anthropic Sonnet 4.5) for better code generation, as NA10's underlying model is not disclosed.
• Node-by-Node Construction and Debugging:
  1. Daily Trigger Node: Built in "10 seconds" using a short, optimized prompt from Vectal.
  2. Set Node: Configured to set two variables, connected by the AI. The speaker praises this node-by-node approach for avoiding "debugging hell" by allowing the NA10 AI to focus on getting each node right.
  3. Google Calendar Node (Get Today's Calendar Event): This more complex node required selecting the calendar.
  4. Enhanced Google Credential Integration: A major improvement in the update is simplified Google credential setup. It now involves a "one-button" process after a "costly Google audit," eliminating the need for Google Cloud Console, client IDs, or OAuth configurations.
  5. Code Node (Transform Events): Vectal provided optimized JavaScript code (31 lines vs. NA10 AI's 42 lines), which the speaker preferred due to Sonnet 4.5's coding capabilities. The NA10 AI correctly identified items as input.all for input.
  6. AI Agent Node (Analyze Calendar and Generate Report): This node uses an external AI model (defaulting to GPT-4.0, but the speaker prefers GPT-4.1 for its reliability and cost-effectiveness). The NA10 AI generated both system and user prompts. The speaker manually added XML tags (<events>...</events>) for better context engineering and adjusted temperature to 0.3 and max tokens to 4,000 (as suggested by Vectal) to ensure conciseness and determinism.
  7. Gmail Node (Send Report via Gmail): The final step, where the NA10 AI automatically connected the Gmail credentials. The speaker manually entered his email address for the recipient and confirmed the subject and message content.
• Final Execution and Debugging:
  • The initial full workflow run resulted in an empty report.
  • Debugging revealed that the "JSON events" output was object object instead of actual data.
  • The NA10 AI, when prompted to fix the issue, surprisingly removed the Gmail node, which the speaker had to manually add back. This highlights that while great for "zero to one" creation, caution is needed for complex, existing workflows.
  • After further adjustments and re-execution, the agent successfully sent a detailed daily calendar report, including total events, average duration, and analysis of each event (type, attendees, duration).

---

Demonstration 3: Long-Form to Short-Form Video Factory Agent

The final demonstration attempts to build the most complex agent: a "long into short form factory" that takes a YouTube video, creates, and publishes multiple shorts with A/B testing.

• Pushing the Limits: The speaker tried to "oneshot" this complex build with a single prompt, which exceeded the 1,000-character limit even after trimming. This reinforced the need for the limit to be increased.
• Complex Output: Despite the prompt limit, the NA10 AI generated a workflow with "25 nodes," a task that would typically take an experienced NA10 user "20-30 minutes" just to assemble, or "hours" for a beginner.
• The "Human-in-the-Loop" Paradigm: The NA10 AI provided a "solid outline" and even listed steps for human intervention (e.g., "get API key," "log into your account") for tasks it couldn't automate. This exemplifies the "human plus AI" model, where humans provide "ideas," "initiative," and "creative vision," while AI handles "repetitive and manual tasks." The speaker predicts this collaboration will be the norm for the "next five to seven years."

---

The Transformative Impact and Future of AI Agents

The speaker concludes by emphasizing the revolutionary impact of NA10's workflow builder:

• Accessibility and Efficiency: The reduced friction and time commitment (from hours to "5 to 10 minutes") make automating even minor inefficiencies worthwhile for individuals and businesses.
• The "Human + AI" Future: The future, especially in the next "6 to 12 months," will see AI handling all repetitive tasks, freeing humans for creative and strategic roles. The "human plus AI" model is crucial, with humans driving vision and agents executing manual work.
• Call to Action: The speaker urges viewers to "start building, start automating" repetitive and boring processes in their lives and businesses. He warns that those who don't will "not be able to keep up" with those who leverage AI agents, as "technology always wins" and is the "ultimate leverage." He encourages pushing through the initial learning curve of the first "one or two agents," as "everything after that is so much easier."

---

Conclusion

NA10's new workflow builder, powered by AI, represents a significant leap forward in the accessibility and efficiency of building AI agents. It dramatically reduces the time and effort required to automate tasks, making it feasible for individuals and businesses to automate even minor inefficiencies. While still in beta with some limitations (like the 1,000-character prompt limit), its ability to generate complex workflows from natural language prompts, pre-fill credentials, and even suggest human intervention steps, positions NA10 as a leader in the automation space. The video strongly advocates for embracing this "human + AI" future, emphasizing that leveraging AI agents is no longer optional but essential for staying competitive and productive.