In today’s fast-paced world, the rate of change is nothing short of astonishing. For companies like Konamiya, keeping up with the latest tools, strategies, and breakthroughs in artificial intelligence can be more challenging than ever. To ensure we always provide our clients with the best advice and the most efficient solutions, we needed a way to continuously monitor, evaluate, and integrate the most critical updates from the AI landscape.

Embracing a No-Code, API-Driven Approach

Rather than writing custom code or maintaining our own servers, we opted for a no-code solution that leverages powerful APIs. This approach enables us to collect, process, and distribute vital information quickly and reliably—all while keeping our operational overhead low. Our primary goal was to build a newsletter that aggregates insights from numerous YouTube channels known for distilling complex AI information into accessible content.

Our Technology Stack

To build this innovative solution, we carefully selected tools that are reliable, cost-effective, and easy to integrate:

• Make.com: The backbone of our system, used to construct the core automation logic.
• RapidAPI: Employed to extract audio from YouTube videos in MP3 format.
• Runpod: Utilized to convert audio to text via a specialized transcription model.
  Affiliate Program: Runpod
• Elevenlabs: Responsible for creating the voice version of the newsletter.
  Affiliate Program: Elevenlabs

Our Data Organization: The Spreadsheet

For our database, we rely on a flexible and straightforward spreadsheet that organizes all the key information required for our newsletter. Our primary spreadsheet contains eight columns:

• STATUS: Reserved for future use.
• ACTION: Reserved for future use.
• IsEnabled: Indicates whether a channel should be monitored.
• Category: Provides context about the channel and helps tailor prompts.
• Channel URL: Direct link to the YouTube channel for quick reference.
• RSS ID: Identifier used by the RSS Feed module in Make.com.
• Group Name: Specifies which newsletter group the channel belongs to (multiple groups are possible).
• Filter Prompt: Used to filter videos based on their title to ensure relevance for the newsletter.

How the Automation Works: A Step-by-Step Overview

Our automated newsletter creation is broken down into several key scenarios within Make.com:

1. YouTube Feed Aggregation

Once a day, at a scheduled time, the system triggers the first phase of automation. The spreadsheet module scans the database for all rows marked as “enabled.” After filtering out irrelevant entries, qualifying records are transferred to a secondary working sheet dedicated to that day’s newsletter. The process then aggregates all the entries and waits for 10 minutes—ample time to ensure that all operations have been completed before proceeding to the next step.

2. Video Processing via RapidAPI

In the next phase, when triggered with the appropriate parameters, the system sends the video URL to RapidAPI. The API fetches the audio in MP3 format, and the corresponding status is updated in the spreadsheet.

3. Handling Transcription Tasks

After processing, RapidAPI returns task information via a callback URL. If successful, the task is forwarded to Runpod for transcription. In cases of error, the system automatically retries the request. Should errors persist, an alternative RapidAPI call retrieves the MP3, and once the task succeeds, transcription via Runpod is initiated.

4. Compiling the Final Newsletter

Triggered by the main scenario after the 12-minute wait, this phase gathers all the data from the newsletter spreadsheet. It performs several tasks:

• Verification: Confirms that videos are not YouTube Shorts.
• Summarization: Sends the transcription to ChatGPT, which generates a concise summary of the video.
• Voice Transcript Preparation: Prepares a text version for the voice newsletter and sends it to ChatGPT.
• Audio Generation: The text is then sent to Elevenlabs, which creates a voice file that is saved to Google Drive.
• Final Assembly: All relevant information—including thumbnails, video links, and summaries—is compiled into a final HTML document via ChatGPT.
• Distribution: The final HTML is emailed through Gmail to the team members who rely on this solution.

Daily Efficiency and Added Value

Our system is designed for continuous updates. Every morning, users can catch up on any content missed from the previous day or new updates released overnight while they were asleep. The voice version of the newsletter, in particular, is perfect for absorbing information during a morning coffee—making it an ideal solution for busy professionals.

Future Enhancements

Looking ahead, we are planning to expand our monitoring capabilities beyond YouTube. Soon, we will include blogs and manufacturer websites to gather additional information, allowing us to overcome the current limitations of relying solely on YouTube content.

Cost Analysis

Creating a newsletter based on 10 videos typically involves about 400 operations in Make.com. The expenses break down as follows:

• GPT API: Approximately 10 cents per newsletter.
• Runpod Transcription: Roughly 1 cent per video transcription.
• Elevenlabs: The most significant cost, with a $22 monthly subscription allowing for about 30 voice newsletters.

Conclusion

By harnessing the power of no-code platforms and a well-integrated technology stack, we have developed an automated, efficient, and cost-effective AI newsletter solution. This system not only keeps us up-to-date with the latest advancements in the AI industry but also ensures that our clients receive timely, high-quality insights in both text and audio formats.

In summary, our automated AI newsletter is a testament to how modern technology can simplify complex processes. With this innovative solution, you can catch up on missed content, absorb updates during your morning coffee, and stay ahead in the dynamic world of AI. Whether it’s through written insights or engaging voice summaries, our goal is to empower you with the information you need to adapt and thrive in an ever-changing digital landscape.

Looking ahead, we are expanding our monitoring capabilities beyond YouTube by including blogs and manufacturer websites, ensuring a richer and more comprehensive flow of information. This means that every morning, you’ll have the opportunity to catch up on any updates from the previous day or overnight developments while you enjoy your coffee.

For further insights into our approach and the world of AI automation, explore our related articles:

• Getting Started with Make.com Automation
• How We Automated Lead Qualification with Retell AI, Make.com, and ClickUp
• How to Kickstart AI Automation in Your Company: A Long-Term Cultural Shift, Not Just a One-Off Project

Let’s Collaborate

If you’re looking for this or any other AI-driven solution to optimize your information flow, streamline your processes, or simply stay ahead in the AI landscape, we’re here to help. Please write to us at hello@konamiya.com and let’s bring your AI solutions to the next level!

At Konamiya, we believe that AI should enhance human capabilities, not replace them. Our vision for AI-driven automation is to simplify, accelerate, and optimize workflows, allowing professionals to focus on high-value tasks.

Previously, we discussed how we automated email drafting for sales teams using Make.com, ClickUp, and OpenAI. Now, we’re introducing a new functionality: automated lead qualification through a Voice AI Agent. This agent follows a structured conversation flow, analyzes the interaction, and logs the insights directly into ClickUp.

The result?

A single agent can multiply their efficiency by prioritizing only the most engaged leads, significantly improving both productivity and service quality for our clients.

Below, we’ll explain how we leverage Retell AI to handle both outbound and inbound calls efficiently.

---

What Is Retell AI?

Retell AI is a cutting-edge platform that enables the creation of AI voice agents for handling phone calls and automating tasks, such as appointment scheduling, with a latency of 600-800 ms. This ensures near real-time responsiveness while balancing computational complexity to maintain a natural conversation flow.

Key Features of Retell AI:

✅ Call transfer between agents for smooth conversation flow
✅ Sentiment analysis to gauge customer emotions
✅ CRM system integration for efficient data management

Designed with security in mind, Retell AI is SOC 2 and HIPAA certified, ensuring compliance with industry standards. The platform is widely used across healthcare, finance, and customer service, making it a reliable choice for businesses looking to enhance their voice communication automation.

---

 

Why We Chose Retell AI Over Other Platforms

Before implementing Retell AI, we thoroughly tested the market for the best solution. We explored several alternatives, including BlendaAI, VoiceFlow, Synthflow, and more. The two finalists were Vapi and Retell AI—both offering similar UI, pay-as-you-go pricing models, and API integrations.

After rigorous testing, Retell AI emerged as the winner, primarily for two reasons:

1. SIP Trunking Support – SIP Trunking significantly reduces call costs by allowing businesses to integrate their own telephony provider. This ensures lower per-minute rates, better call quality, and scalable infrastructure to handle high-volume calls efficiently.
2. Built-in Free Agent Testing – Retell AI allows both chat and voice testing directly in its dashboard, making development and debugging much smoother.

Additionally, Retell AI offers seamless integration with Cal.com, enabling effortless meeting scheduling. With just a few configuration tweaks, we automated the booking process, further streamlining our lead management pipeline.

---

 

Our Automated Call Workflow with Retell AI

We handle both outbound and inbound calls, with a structured flow designed to maximize efficiency.

1. Outbound Calls: How It Works

Our AI-powered lead qualification process starts when an agent selects “Schedule AI Lead Phone Call” from the action menu in ClickUp.

Step-by-Step Flow

1️⃣ ClickUp triggers an automation, sending a webhook request to Make.com.
2️⃣ Make.com verifies the data:

• Checks if a phone number exists.
• Validates the phone number format.
• Retrieves lead details.
• Extracts existing ClickUp comments to understand prior interactions.
• Analyzes data via OpenAI to personalize the call script.
  3️⃣ The call is registered in Retell AI, and ClickUp statuses update accordingly (e.g., “CALL REGISTERED” or “ERROR”).
  4️⃣ Retell AI executes the call and streams real-time event updates back to Make.com.

Post-Call Processing:

✅ Retell AI sends the transcript and call outcome.
✅ OpenAI analyzes the transcript and generates a structured JSON report.
✅ Additional data from Retell AI is appended for more comprehensive insights.
✅ The final summary is logged as a ClickUp comment, ensuring the agent has clear, actionable insights.

---

2. Handling Inbound Calls

Inbound calls are handled differently. Since we focus on outbound lead engagement, we only accept calls from leads we’ve already contacted.

Step-by-Step Flow

📞 When a lead calls back, Retell AI sends a webhook request to Make.com.
🔍 Make.com cross-references the number with outbound call logs.
📂 The system retrieves relevant data about the previous interaction.
🗣️ Retell AI receives an updated conversation context, allowing the AI agent to respond intelligently.
✅ Post-call processing follows the same structured flow as outbound calls, ensuring data consistency.

---

Conclusion: Key Takeaways & Compliance Considerations

Retell AI provides an advanced, flexible, and cost-effective voice automation solution, with extensive customization through its dashboard, API, and SIP trunking capabilities.

By integrating Retell AI, Make.com, and ClickUp, we have successfully automated lead qualification, enabling sales teams to focus on the most promising prospects—without wasting time on unqualified leads.

🚀 More automation, better efficiency, and higher sales conversions.

---

Compliance Considerations for AI-Powered Calls in Europe

When conducting AI-driven voice calls, especially within the European Economic Area (EEA), it is crucial to adhere to GDPR (General Data Protection Regulation) and local telecommunications laws.

Key Best Practices for Compliance:

📌 Consent for Call Recording – If calls are being recorded, you must explicitly inform the lead at the beginning of the call and, in some cases, obtain their consent. A typical disclaimer could be:
“This call is being recorded for quality and compliance purposes. If you do not wish to be recorded, please let us know.”

📌 Legitimate Interest vs. Consent – Under GDPR, processing personal data (including recorded conversations) requires a lawful basis. If call recording is not strictly necessary, obtaining explicit consent is the safest approach.

📌 Right to Be Forgotten – Leads must have the right to request deletion of their data, including call recordings. Your system should provide a mechanism for handling such requests.

📌 AI Disclosure Requirement – Under GDPR, businesses must inform users when they are interacting with an AI-driven system. A best practice is to offer an option to speak with a human agent when requested.

📌 Respecting “Do Not Call” Lists – Some EU countries maintain opt-out databases for unsolicited calls. Ensure compliance by verifying leads against such lists before initiating outbound calls.

📌 Data Storage & Security – If call transcripts or recordings are stored, they must be protected with strong encryption and access controls, ensuring only authorized personnel can access them.

📌 Data Retention Policy – Companies should define a clear data retention policy for call recordings and transcripts, ensuring compliance with GDPR data minimization principles.

---

🚀 Ready to Automate Your Lead Qualification Process?

Want to see Retell AI in action? Contact us today to explore how AI-driven voice automation can improve your lead qualification process and help your team convert more leads with less effort! 🚀

Artificial Intelligence has been part of our world for many years. Initially, it manifested through models that categorized our text or images, or through algorithms that intelligently matched our needs based on what we typed in search engines. It helped us find the best routes, optimized advertising budgets, and made our lives more efficient—often without us even noticing.

Yet, on November 30, 2022, everything changed. The public release of ChatGPT created a seismic shift in how we view AI. Visionaries worldwide lit up with fresh ideas and novel use cases. Early systems based on ChatGPT started simply—integrating with PDFs, CSVs, and booking services—but that was just the beginning. Soon enough, more advanced tools like Perplexity combined ChatGPT’s language abilities with real-time data scraping, while other companies (such as Anthropic) launched specialized models. In under three years, we’ve seen entire text editors (like Cursor) reshape the way we code and build applications.

Stepping into the Realm of AI Agencies

At Konamiya, we believe the next leap forward lies in AI Agencies: autonomous systems with defined goals and the ability to adapt to changing environments. These agencies:

• Operate based on language models, not static code.
• Continuously improve with each new model release, lowering development overhead.
• Allow real-time monitoring and interaction, letting you issue new tasks and observe their decisions.

In other words, they become digital coworkers: systems that not only handle complex tasks but also learn and grow alongside your business, helping to reduce costs and streamline daily workflows.

Why Host OpenAI Agents on Discord?

One of the most user-friendly platforms to host and interact with AI agents is Discord. Known primarily for its chat rooms and community features, Discord also offers:

• Real-time Communication: Multiple channels dedicated to different AI agents and tasks.
• Accessibility: Team members can instantly view agent outputs and provide feedback.
• Cost Optimization: Hosting agents on a lightweight platform like Discord can mitigate infrastructure overhead and centralize monitoring, cutting down on unnecessary spending.

With Discord, managing AI agencies becomes less daunting. You can keep an eye on everything—whether it’s a support bot answering questions or a creative content generator producing fresh ideas—without juggling multiple disconnected tools.

Agency Swarm and Other AI Frameworks

In this new era of autonomous AI, several frameworks have emerged to help you orchestrate multiple agents:

• Agency Swarm: Offers flexible orchestration, customizable agent roles, and efficient communication among agents.
• Copilot Studio: Integrates deeply with Microsoft 365 and allows low-code/no-code development of AI assistants.
• CrewAI: Focuses on a team-based approach where multiple AI agents collaborate under defined roles.
• Autogen, LangGraph, Langflow, and Vertex AI: Each introduces unique architectures—some with drag-and-drop interfaces, others with advanced multi-agent workflows.

While each solution offers unique features, a major advantage lies in combining these frameworks with Discord to achieve lower operational costs, high availability and reliability, as well as robust security—all without incurring the overhead typically associated with more complex infrastructures.

A Vision for the Future

We stand on the threshold of an unprecedented AI-driven landscape. As these agencies become more advanced, they won’t just chat with users—they’ll make decisions, handle multifaceted tasks, and scale processes across entire organizations. By tying their operations to continually improving language models, businesses can reduce coding costs, simplify maintenance, and rapidly adapt to shifting market demands.

Imagine a future where:

• Customer service is instantly responsive, aided by specialized agents that proactively handle routine issues.
• Development teams rely on AI for rapid prototyping, code suggestions, and automated testing.
• Marketing campaigns become fully automated—from creating tailored content to analyzing performance in real time.

This isn’t just wishful thinking. It’s the natural evolution of AI as we shift from single-use chatbots to autonomous swarms of agents capable of collaborating, learning, and driving innovation at scale. And thanks to platforms like Discord, the barrier to entry is lower than ever—helping organizations optimize costs while gaining a powerful competitive edge.

Conclusion

By hosting OpenAI Agents on Discord and leveraging tools like Agency Swarm, even small teams with limited budgets can have a centralized hub where all communication with their AI agents takes place. This solution is not only cost-effective and straightforward to implement using readily available AI services, but it also scales seamlessly as the number of agents grows. In turn, organizations gain a flexible system that’s both easy to update and adaptable to evolving needs, all while keeping operational overhead to a minimum.

For businesses eager to stay ahead, the path is clear: Embrace AI agencies, invest in lightweight yet powerful platforms, and watch as digital collaborators transform every facet of your operations. The future is here, and those who act now will shape its course.

Agency Swarm is an advanced framework for orchestrating multiple AI agents, built on top of the OpenAI Assistants API. Originally created by Arsenii Shatokhin (VRSEN), its primary goal is to support fully automated AI agencies, providing a flexible, powerful toolset for managing a variety of tasks across different domains.

---

1. Introduction to Agency Swarm

In the rapidly evolving world of AI, businesses and developers are seeking more efficient and scalable solutions for automating operations. Agency Swarm emerges as a response to this need:

• It coordinates multiple AI agents simultaneously, allowing them to communicate and collaborate.
• It features a role-based system, enabling you to define each agent’s responsibilities.
• It offers lightweight yet effective state management, making it suitable for production environments.

---

2. System Architecture

2.1 Key Components

• Customizable Agent Roles: Assign responsibilities and tasks to each agent based on your specific requirements.
• Inter-Agent Communication Mechanism: Facilitates streamlined dialogue among agents, without forcing a rigid hierarchy.
• State Management Tools: Uses settings.json to track the status of assistants, keeping each task contextually relevant.
• Handoff System: Allows tasks to be smoothly passed from one agent to another.

2.2 State Management

• The framework stores assistant states in a JSON file to ensure consistency.
• Tasks run in a stateless manner, significantly reducing memory overhead.
• Context is managed efficiently, allowing each agent to stay focused on relevant data while avoiding confusion or “hallucinations.”

---

3. Unique Features

3.1 Communication System

• Relies on a dedicated SendMessage tool for agent orchestration.
• Unified, non-hierarchical communication structure means agents can collaborate more organically.
• Customize your own communication flows—ideal for complex workflows involving multiple tasks and stakeholders.

3.2 Asynchronous Modes

• async_mode='threading': Enables asynchronous communication among agents.
• async_mode='tools_threading': Allows parallel execution of agent tools for even faster results.
• Agents can also share files, making it easier to collaborate on large projects.

---

4. Practical Applications

4.1 WebDevCrafters

An agency specializing in Next.js, React, and MUI uses Agency Swarm to automate development tasks—ranging from generating boilerplate code to QA checks—freeing developers to focus on creative problem-solving.

4.2 CodeGuardiansAgency

By integrating Agency Swarm with GitHub Actions, this team automates code reviews, ensuring code consistency and alignment with Standard Operating Procedures (SOP). It also provides thorough documentation, minimizing the risk of missing important details.

---

5. Competitive Advantages

When set against other multi-agent frameworks, Agency Swarm distinguishes itself by:

1. Avoiding Extra Model Calls: Unlike certain competitors, it doesn’t require additional model invocations just to identify which agent should speak next, streamlining communication.
2. Better Task Flow Management: Enjoy greater control over how tasks are distributed among agents.
3. Lower Risk of Hallucination: With robust type validation and fewer dependencies, it minimizes the chances of AI straying from valid data.

Compared to CrewAI

• No LangChain Dependency: Reduces complexity and overhead.
• Stronger Validation: Decreases error rates and confusion.
• Lean Communication Flow: Simplifies how agents interact, making adoption easier for smaller teams.

---

6. Advanced Functionalities

6.1 Execution Control

• Configure temperature and token parameters to tailor each agent’s output.
• Apply truncation strategies to prevent runaway tasks.
• Fine-tune individual agents for specialized roles, such as drafting blog posts or performing data analysis.

6.2 Developer Tools

• Integration with Instructor for type validation, reducing inconsistencies.
• Automatic error correction mechanisms, catching issues before they escalate.
• Extend the framework with custom tools—perfect for building domain-specific capabilities or connecting to external APIs.

---

7. Business Benefits

7.1 Operational Efficiency

• Faster task execution saves valuable time for your team.
• Intelligent resource allocation means agents focus on the most critical tasks first.
• Mundane processes become automated, freeing human resources for more innovative work.

7.2 Scalability

• Easily add new agents as your needs grow, without overhauling the entire system.
• Flexible architecture adapts to changing workflows and business goals.
• Straightforward deployment in production—ideal for companies looking to scale quickly.

---

8. Challenges and Limitations

8.1 Technical Hurdles

• Requires a basic grasp of programming to configure properly.
• Official documentation can be sparse, leading to a steeper learning curve.
• Initial setup complexity might deter teams new to multi-agent frameworks.

8.2 Operational Considerations

• Dependence on OpenAI’s API means you’ll need to track usage and costs.
• Ongoing monitoring and agent management is crucial, especially as tasks increase.
• Potential API cost spikes if agents perform large-scale or frequent tasks.

---

9. The Future of Agency Swarm

Development of Agency Swarm shows no signs of slowing down. Upcoming enhancements include:

• Additional integrations with leading AI and workflow platforms.
• Improved agent management interfaces to reduce complexity.
• Extended automation features, pushing the boundaries of what multi-agent systems can achieve in real-world business contexts.

---

Conclusion

Agency Swarm stands as a powerful, flexible framework for orchestrating multiple AI agents in a seamless, highly configurable environment. By providing advanced communication mechanisms, role-based architecture, and accessible state management, it empowers businesses to automate tasks, streamline operations, and stay competitive in an increasingly AI-driven market.

For organizations searching for a robust multi-agent solution that maximizes efficiency while scaling easily, Agency Swarm offers a dynamic path forward—one that continues to evolve in tandem with cutting-edge AI advancements.

In the realm of complex marketing and sales processes, Google Apps Script can prove indispensable. By leveraging it, you can automate numerous tasks in Google Sheets, manage workflows in ClickUp, integrate with tools like Make.com, and ultimately save your team valuable time.

---

Why Google Apps Script?

Simplicity and flexibility — these two qualities make Apps Script stand out from other solutions. It runs directly within the Google ecosystem, eliminating the need to build and maintain a separate infrastructure. With a comprehensive range of built-in methods and the ability to connect to external APIs, Apps Script excels at:

• Secure web interface
• Automating marketing activities (e.g., lead scoring)
• Managing data in spreadsheets
• Creating notifications and scheduling tasks
• Streamlining communication among multiple tools

---

Key Use Cases and Trigger Types

Many people are unaware that Google Apps Script offers various trigger types to run scripts on demand. This means you can respond to specific spreadsheet events or schedule tasks at fixed times.

1. Time-Driven Triggers

By setting time-based schedules (e.g., every 10 minutes, daily, or weekly), you can:

• Update data in spreadsheets
• Send notifications to your team
• Generate reports and email them automatically
• Check if something happens every X time

2. Change-Driven Triggers

• onEdit

  • Fires only when a user changes a cell value via the browser interface.
  • Runs automatically as a simple trigger.
  • Does not detect adding rows, columns, or comments.

  It’s perfect for scenarios where you want to run an Apps Script function based on user-driven actions—for instance:

  • Updating records when a cell in an “Action” or “Status” column is changed by the user, allowing specific rows to trigger Apps Script functions on demand

  ---

  onChange

  • Requires manual installation (it’s an installable trigger).
  • Can detect various types of changes, such as:
    • EDIT (cell edit)
    • INSERT_ROW (adding a row)
    • INSERT_COLUMN (adding a column)
    • REMOVE_ROW (removing a row)
    • REMOVE_COLUMN (removing a column)
    • FORMAT (formatting changes)
    • OTHER (other types of modifications)

  It’s particularly useful when you need to automatically run actions after a scenario completes in Make.com or when a third-party service updates your spreadsheet.

3. “onOpen” Triggers

When a user opens a spreadsheet, you can automatically:

• Show a sidebar with important instructions
• Refresh auxiliary data or metrics
• Verify user permissions
• Perform functions
• etc…

---

Lead Scoring in Action

A prime example of how Google Apps Script shines is in lead scoring. It typically involves:

1. Adding a new lead to a Google Sheets spreadsheet
2. Running a script that sends the data to an analytics tool (e.g., an AI model) to generate a lead score
3. Saving that score back into the spreadsheet and forwarding key details to other systems — such as a CRM or marketing automation tool

By combining both time-driven and change-driven triggers, the process remains continuously monitored, and any errors can be handled via automated retry mechanisms.

---

Connecting to the Broader Ecosystem

An organization like Konamiya doesn’t limit itself to Google-only tools. Its ecosystem might include:

• ClickUp – for project and task management
• Make.com – for orchestrating workflows among different platforms
• Other tools dedicated to automating contact management and relationship building

Apps Script acts as the “glue” linking Google Sheets to these systems, enabling smooth data exchange. For instance, when a new lead is added to the spreadsheet, a script can call Make.com, which triggers a task in ClickUp and initiates message sending in an automation tool.

---

Security and Reliability

Any automated process demands a strong focus on security and stability. With Google Apps Script:

• You can restrict access to specific spreadsheets and scripts to prevent unauthorized edits
• You can configure retry mechanisms if a script encounters an error
• A variety of triggers ensure that essential tasks (e.g., updating lead databases) aren’t missed, even in the event of intermittent network or API issues

---

Why It’s Worth It

Harnessing Google Apps Script doesn’t just streamline tasks like lead scoring, reporting, and communications—it also provides a user-friendly, graphical interface that can accommodate a vast range of integrations and functionalities. In many ways, your only real limit is your own imagination. For organizations that:

• Operate within Google’s ecosystem and want to automate specific tasks,
• Need seamless integration with external services (CRMs, marketing tools, or other platforms),
• Prioritize ease of deployment and low maintenance costs but also desire versatile customization,

Apps Script truly shines as a winning solution. With minimal overhead and a robust set of built-in features, it offers unparalleled flexibility for developers and non-developers alike—elevating productivity without requiring extensive infrastructure.

---

Questions or looking for more info?
Get in touch. We’re always happy to share our experience and help you set up your own automation!

 

 

 

When there’s a need, there’s a solution. Preferably a simple one. This motto guides me whenever Hubert suggests new functionalities for our tools. Recently, we faced a challenge in one of our sales funnels where the high volume of emails required a more efficient flow. Hubert proposed a solution: automatically drafting emails to save time and streamline the process.

From Problem to Vision: Designing a Scalable Workflow

Hearing this need, I immediately began brainstorming solutions. In moments like this, I focus on ideas that are:

1. Easy to implement and maintain,
2. Highly scalable,
3. Reliable.

While cost and execution time are important, they take a backseat to these priorities.

System Components: The Building Blocks of Automation

To design the solution, I started by analyzing the components we had:

• Landing Page
• Spreadsheet (serving as a database)
• AppScript
• ClickUp CRM
• Google Workspace (Gmail/Email)

Defining the Interface and Action Flow

The next question was: What will the interface look like, and how will this function be triggered?

We didn’t want to automatically generate drafts for all emails—this would be unnecessary and problematic. On the other hand, we needed visibility into what happens and when.

With these considerations, I envisioned a system where:

• The action is triggered directly in ClickUp via predefined custom fields in the task template.
• A custom automation in ClickUp sends data to a webhook when the “CREATE EMAIL DRAFT” action is selected.
• The webhook triggers a Make.com scenario that processes the request.

---

The Workflow: Simplified and Powerful

Step 1: ClickUp Triggers the Action

• Using ClickUp’s custom fields, users can select the “CREATE EMAIL DRAFT” option in the task template.
  

Step 2: Clickup to Make.com(Webhook)

The Clickup automation sends data to a Make.com scenario. Here, the entire logic of the functionality is defined.

Assistant Generates the Draft

From Make.com, specific data is retrieved from the database and sent to the OpenAI assistant. Using a carefully crafted prompt and configuration, OpenAI generates:

• The email subject
• The email body

Step 3: Draft Creation and Status Update

The generated draft is created in Gmail, and the task status is automatically updated in ClickUp to reflect the action.

---

Why Make.com? Flexibility, Scalability, and Security

One of the reasons we relied on Make.com for this solution is its flexibility.

• Adding or modifying steps is simple.
• Integration with other components (e.g., Google Workspace) is seamless.
• Maintenance is easier compared to hosting a conventional application on a private server or cloud service.

Security Advantages

Make.com also simplifies security management:

• Webhooks can be restricted to specific IP addresses.
• You can add a unique token to every request, which is validated at the start of the scenario.

---

Continuous Data Synchronization with AppScript

Another vital component of the system is AppScript, which monitors Gmail inboxes to ensure the database remains up-to-date. This ensures that email drafts always contain the latest information and agreements from all relevant sources.

 

Future Enhancements

This solution is designed to be scalable. While this article focused on the email drafting workflow, future updates will dive deeper into how AppScript powers real-time data synchronization. Stay tuned! 😊