Speech Support Workflow
Configured with Agent, LM Chat Google Gemini, Memory Buffer Window, Memory Manager, and additional nodes. Review the connections and setup before adapting this template.
Node categories in this workflow
- AI & models
- Trigger
- Transform
- Messaging
- Routing
- Code
- Data store
Icons describe general node roles, inferred from node types. Inspect the parameters for the configured action.
Drag to pan; use + and − to zoom. Reconstructed from the exported node positions; nothing runs here. Dashed lines represent non-main connections. Full node names and notes appear below.
How it works
The exported workflow contains the following configured nodes. Their connections are shown above; inspect the JSON for the complete logic, expressions, and external actions.
| Node | Category | Type | Operation |
|---|---|---|---|
| AI Agent | AI & models | Agent | See node parameters |
| Google Gemini Chat Model | AI & models | LM Chat Google Gemini | See node parameters |
| Recieve Telegram Message | Trigger | Telegram Trigger | See node parameters |
| Check For Text or Voice Message | Transform | Edit Fields | See node parameters |
| Download Audio File | Messaging | Telegram | file |
| Transcribe Audio File | AI & models | Open Ai | audio / transcribe |
| If Voice Message | Routing | If | See node parameters |
| Code to remove unwanted characters from LLM response | Code | Code | See node parameters |
| Code to split output into chunks under 4000 characters | Code | Code | See node parameters |
| Respond to Telegram Message | Messaging | Telegram | See node parameters |
| Wipe Conversation Memory | Data store | Memory Manager | See node parameters |
| Store Conversation Memory | Data store | Memory Buffer Window | See node parameters |
| Set prompt to start a new speech | Transform | Edit Fields | See node parameters |
| Set prompt to generate a speech based on the feedback | Transform | Edit Fields | See node parameters |
| Set prompt to provide feedback on speech | Transform | Edit Fields | See node parameters |
| Route Flow Based on Message Content | Routing | Switch | See node parameters |
Setup
- Download the JSON and import it into an isolated n8n instance. Leave the workflow inactive.
- Connect your own credentials through n8n’s credential store. Replace
REPLACE_WITH_*values with your resource identifiers. - Inspect API destinations, Code nodes, database queries, file paths, and outbound actions.
- Run with synthetic data and harmless destinations. Enable triggers only after checking the actual outputs.
Requirements
An n8n instance with the listed node types and compatible versions. Service-connected workflows may require accounts, credentials, or paid APIs.
Credential types declared in the source
googlePalmApiopenAiApitelegramApi
Workflow notes
Dynamic System Prompting:
Dynamic System Prompting:
This node sets the AI's system prompt according to the user's request identified in the incoming message.
Telegram-Ready Output: Formatting and Length Management:
Telegram-Ready Output: Formatting and Length Management:
These code nodes perform two crucial tasks:
- Formatting: Removing characters that could cause issues with Telegram's message parsing.
- Chunking: Dividing messages longer than Telegram's 4000-character limit into multiple shorter messages for sequential delivery.
Clearing AI Agent Memory:
Clearing AI Agent Memory:
This node clears the AI agent's short-term memory. This helps to minimize the influence of past interactions on future responses, thereby reducing the likelihood of the AI generating inaccurate or irrelevant information (hallucinations).
Processing Telegram Messages:
Processing Telegram Messages:
This section handles incoming messages from Telegram. It first checks if the message contains text.
- Text Message: If the message includes text, it's directly routed to the analysis switch node.
- Audio Message: If the message is an audio file:
- The audio file is downloaded.
- The audio is transcribed into text.
- The transcribed text is then sent to the analysis switch node.
Finally, the analyzed text (whether directly from a text message or transcribed from audio) is forwarded for further processing based on the analysis results.
Gemini-Powered Response and Conversation Storage:
Gemini-Powered Response and Conversation Storage:
This node utilizes the Google Gemini model to generate a response to the user's prompt and stores the ongoing conversation.
Speech Support Workflow
Speech Support Workflow
Purpose and structure
This folder contains the Speech Support Workflow workflow. Its executable nodes use agent, code, if, lmChatGoogleGemini, memoryBufferWindow, memoryManager, openAi, set, switch, telegram, telegramTrigger. The table and connection map below are derived from the JSON, not an inferred guarantee of behavior.
Setup
Import inactive. Configure your own credentials and replace REPLACE_WITH_* values. Review external destinations and Code nodes before running.
Entry points
Recieve Telegram Message
Verification
Sanitized source candidate; not runtime-tested. This overview is generated from node metadata. Read the folder README for the connection map, credential types, and license.
Attribution
https://github.com/Zie619/n8n-workflows/blob/94007c1445d9258a7da116646b79473e7c7c3282/workflows/Telegram/1452_Telegram_Stickynote_Automate_Triggered.json MIT; preserve source notices.
Canvas section labels
These name areas of the original canvas; they do not include explanatory instructions.
- This n8n workflow acts as your personal AI speechwriting coach, directly accessible through Telegram. It listens to your spoken or typed drafts, provides insightful feedback on clarity, engagement, structure, and content, and iteratively refines your message based on your updates. Once you're ready, it synthesizes a brand-new speech or talk incorporating all the improvements and your accumulated ideas. This tool streamlines the speechwriting process, offering on-demand AI assistance to help you craft impactful and well-structured presentations.