Budget: 5600 PLN Deadline: 14 days
Hello! The task of creating a Telegram bot for working with neural networks is completely clear to me. I specialize in developing complex automation systems and integrating AI models (over 30 successful cases). Since you are looking for a partner for long-term collaboration, my approach is based on creating a scalable architecture rather than just a one-time script. Below is my vision for implementation and technical details.
🧠 My vision for the project implementation
To ensure the bot can handle the load and is user-friendly, I propose the following structure:
1. Technology stack
* Language: Python (aiogram library 3.x). This is the modern standard for asynchronous bots with high load.
* Database: PostgreSQL (for stable storage of user profiles, balances, and request history).
* Request queues (Celery + Redis): Essential for AI bots. Since neural networks generate responses in 5 to 30 seconds, queues will allow the bot to avoid "hanging" and process hundreds of requests simultaneously.
2. Integration of neural networks (Multi-AI)
I will implement a universal connector that will allow connecting:
* Texts: GPT-4o, Claude 3.5 Sonnet, or local models (Llama 3).
* Images: Midjourney (via API), Stable Diffusion, or DALL-E 3.
* Voice: Whisper for transcription and TTS for voice output.
3. Functionality and Monetization
* Token/balance system: Flexible configuration of request costs for each model.
* Personal account: History of dialogues, selection of the "role" of the neural network (copywriter, programmer, translator).
* Admin panel: Real-time statistics, user management, and newsletters.
* Payments: Integration of crypto acquiring (CryptoPay) and bank cards (Stripe/Mono).
⏱ Terms and cost (MVP version)
The estimate is based on creating a stable product ready for market launch:
* Execution time: 10–14 working days.
* 3 days: Architecture and integration of main APIs.
* 5 days: Bot logic, personal account, payments, and admin panel.
* 4 days: Load testing and deployment on the server.
* Cost: $800 – $1,500 (the final amount depends on the number of neural networks to be integrated in the first version).
Why me:
I have experience in optimizing API costs (caching identical requests), which will help you save on token budgets. I will set up the system so that adding new features in the future will not require rewriting all the code.
Bonus: I will set up Docker containerization and automatic deployment to your server (Hetzner/DigitalOcean) for free, so the bot runs 24/7 without failures.
Please send the complete technical specification.