Your AI assistant for Telegram

Description

👥 Who Is This Bot For?

🧠 Knowledge Workers & Creators
Summarize ideas, brainstorm on the go, or automate voice-based journaling.

🧰 Workflow Developers & Engineers
Trigger APIs or run n8n automations using natural Telegram messages.

🎧 Customer Support & Helpdesk Teams
Deliver real-time replies with AI that remembers user context.

🗣️ Voice-First & Accessibility Champions
Build inclusive bots that support voice and text equally — using Telegram + Whisper + GPT-4.

❌ What Problems Does It Solve?

⚠️ Voice messages aren’t handled intelligently or automatically
⚠️ Responses are repetitive, with no memory or context
⚠️ Voice input is ignored or unsupported in most bots
⚠️ Users get generic replies due to lack of contextual awareness
⚠️ Building scalable logic often requires heavy coding

✅ How This Bot Solves It

🎙️ Listens to voice or text messages in real time
🧠 Remembers the last 10 messages for contextual awareness
🗣️ Converts voice to text using OpenAI Whisper
🧾 Uses GPT-4 to generate meaningful replies
🖋️ Sends clean, HTML-formatted messages back to Telegram
🔄 Handles unknown input types gracefully with fallback logic

🧭 How It Works — The Workflow

1️⃣ Trigger
Activated on new Telegram messages (text or voice)

2️⃣ Input Detection
Checks message type and routes accordingly

3️⃣ Transcription
Voice input is transcribed using Whisper

4️⃣ Message Assembly
Combines message content and type for AI processing

5️⃣ AI Reply
Sends the assembled message to GPT-4 with memory context

6️⃣ Final Output
Posts a polished reply in Telegram with proper formatting

🛠️ Customization Features

🧠 Memory Options
Adjust context window or toggle memory completely

💡 Model Flexibility
Switch between GPT-4, Claude, Mistral, or any custom LLM

🔊 Speech Engine Choices
Replace Whisper with Deepgram, AssemblyAI, or Azure Speech

📝 Response Format
Choose HTML, Markdown, or plain text formatting

🌐 API Integration
Trigger external tools, CRMs, or n8n workflows directly from replies

🚧 Fallback Handling
Custom error messages and graceful failover for edge cases