AI-Powered Speechwriting Assistant – Workflow

Public speaking terrifies most people more than death itself. The sweaty palms, the racing heart, the blank mind when you step up to the podium. But what if the biggest hurdle wasn't the speaking part, but the writing part?

Crafting a compelling speech takes time, skill, and lots of back-and-forth refinement. You write a draft, read it aloud, cringe at the awkward transitions, rewrite sections, and repeat the cycle until you're either satisfied or out of time.

This n8n workflow changes everything. It creates your personal AI speechwriting coach that lives right in your Telegram app. Send it voice recordings of your practice sessions or text drafts, and get instant, detailed feedback. When you're ready, it synthesizes everything into a polished final speech that incorporates all your improvements.

Prerequisites: What You Need to Get Started

Before diving into this workflow automation magic, you'll need a few API accounts set up:

• Telegram Bot API: Create a bot through BotFather to receive and send messages
• Google Gemini API: Powers both the conversational AI and audio transcription features
• n8n Platform: Either self-hosted or cloud version to run the workflow

The beauty of this setup? Everything happens through Telegram, so you can practice speeches during your commute, get feedback while walking, and refine your content anywhere inspiration strikes.

Key Components: The Building Blocks

This workflow uses several powerful n8n nodes working in harmony:

• Telegram Trigger: Captures incoming messages and voice notes
• Google Gemini Transcription: Converts voice recordings to text
• Switch Node: Routes requests to different AI coaching modes
• AI Agent Nodes: Three specialized coaching assistants
• Memory Manager: Maintains conversation context and learning
• Code Nodes: Format responses for Telegram delivery

Step 1: Set Up Message Processing Pipeline

The workflow starts with a Telegram trigger that listens for your messages. When you send something, it first checks whether you've sent text or voice. This flexibility means you can type quick questions or record yourself delivering speech sections naturally.

If you send a voice message, the Google Gemini transcription node converts your audio into text with impressive accuracy. This transcribed text then joins the same processing pipeline as regular text messages, creating a seamless experience regardless of input method.

Step 2: Intelligent Message Routing

Here's where the workflow gets smart. A switch node analyzes your message content and routes it to one of three specialized coaching modes:

• "New speech" trigger: Starts fresh speech preparation sessions
• "Generate speech" command: Creates complete speeches from your accumulated ideas
• Default mode: Provides detailed feedback on speech drafts

This routing system means you get the right type of help at the right time, whether you're brainstorming ideas, refining content, or ready for final synthesis.

Step 3: Configure Three AI Coaching Personalities

Each coaching mode gets its own AI agent with a specialized system prompt. The preparation coach asks probing questions about your audience, key messages, and speech structure. The feedback coach analyzes clarity, engagement, pacing, and impact with specific suggestions. The synthesis coach combines everything into polished final content.

These aren't generic AI responses. Each agent understands its specific role in the speechwriting process and provides targeted, actionable guidance that builds on your previous interactions.

Step 4: Implement Smart Memory Management

The workflow includes sophisticated memory handling that makes long-term coaching possible. It maintains a 25-message conversation window, so the AI remembers your previous drafts, feedback received, and improvements made.

When you start a new speech project, the memory manager can clear previous conversations to prevent confusion between different speech topics. But within a single project, it builds understanding over time, leading to increasingly personalized and relevant feedback.

Step 5: Format and Deliver Responses

The final step handles the practical side of Telegram delivery. Two code nodes work together to make responses readable and manageable. The first removes formatting characters that could break Telegram's message parsing. The second splits long responses into chunks under 4000 characters.

This means you get comprehensive feedback without hitting Telegram's message limits. Long speech analyses arrive as a series of well-structured messages that you can read and reference easily.

Real-World Benefits and Use Cases

This AI speechwriting coach transforms how you approach public speaking preparation. Instead of struggling alone with drafts, you get instant expert feedback that improves with each interaction.

For Business Professionals: Perfect keynote speeches, presentation pitches, and board room updates during your busy schedule. Practice in your car, get feedback during lunch breaks.

For Students and Academics: Refine thesis presentations, conference talks, and classroom presentations with detailed structural and content feedback.

For Event Speakers: Develop engaging wedding toasts, eulogy speeches, and celebration remarks that truly connect with audiences.

The workflow learns your speaking style and preferences over time, becoming a truly personalized coaching experience. It's like having a professional speech coach available 24/7 through a simple messaging interface.

Most importantly, this automation removes the biggest barrier to great public speaking: the overwhelming complexity of crafting compelling content. With AI handling the heavy lifting of analysis and synthesis, you can focus on what matters most – connecting with your audience and delivering your message with confidence.

Ready to transform your public speaking game? This n8n workflow puts professional-grade speechwriting assistance right in your pocket.