Development of a Telegram userbot with personality and memory (based on Chat GPT)

Technical assignment: Intelligent Userbot with personality and memory (based on ChatGPT)

The goal of the project is to create a scalable userbot based on ChatGPT, capable of conducting personalized, emotionally realistic dialogues on behalf of multiple accounts. The bot must have a stable personality, contextual and global memory, the ability to dynamically manage behavior, and support safe operation on the server.

Main functional requirements:

1. Stable image and personality of the bot

 • Each userbot must have an individual biography, interests, behavioral traits, speech style, and logic of thinking. (Scalability after testing)

 • Behavior and information delivery must be consistent, without contradictions.

 • Ability to set/update the legend separately for each account.

2. Memory and context

 • Short-term memory: storing and using the context of the current dialogue.

 • Global memory: storing key facts, reactions, and client features on a secure server.

 • Ability to refer to long-term information and adapt based on interactions.

3. Working with multiple accounts and chats

 • Support for working with multiple Telegram accounts or other platforms.

 • Conducting multiple chats simultaneously without information overlap.

 • Isolated behavior logic for each chat and account.

4. Storage and security

 • Secure storage of all data: dialogues, metadata, legends, interaction history.

 • Ability to scale storage and flexible data structure.

 • Support for connecting proxies and anonymization mechanisms.

5. Flexible bot management

 • Ability to manually set a goal or behavior scenario for the bot for a specific day.

 • Changing tasks or communication style on the fly (during the dialogue).

 • Integration with the operator system: transferring the dialogue to manual mode by trigger or request.

 • Tools for real-time operational management.

6. Anti-spam logic and natural communication style

 • The bot should avoid mechanical, repetitive phrases. Behavior should be as natural and "human" as possible.

 • Built-in anti-spam system:

 • Tracking frequency, repetitions, template responses.

 • Ability to vary formulations, tone, and style depending on the client.

 • Automatic switch to "live style" if excessive automation is detected.

 • Imitation of natural pauses, reactions, logical transitions.

 • Support for key triggers and behavior scenarios during suspicious activity.

7. Alert and trigger system

 • Customizable list of keywords or phrases (triggers).

 • When a trigger is activated — automatic notification sent to a Telegram channel (or operator bot).

 • Ability to dynamically update the list of triggers and response scenarios.

8. Architecture and scalability

 • Modular and flexible architecture supporting horizontal scaling.

 • Ability to connect external knowledge bases, ready scripts, and logic (will be provided).

 • Ability to integrate with the GPT API, as well as model updates without service disruption.

9. Additional capabilities

 • Logging actions, automatic report generation.

 • Support for an external interface for managing bots and monitoring dialogues.

 • Ability for step-by-step implementation (MVP → full-scale).

Infrastructure requirements:

 • Server with secure data storage.

 • Ability to connect Telegram, Web, and other platforms.

 • Use of OpenAI API (or other LLM platforms) with migration capability.

Additionally provided:

 • Dialogue logic

 • Ready communication scripts

 • Knowledge base and frequently asked questions

 • Trigger dictionary for alerts

Expected result:

 • Personalized personality of each bot: biography, hobbies, habits, unique speech style

 • Global and short-term memory: consistency in dialogue, without loss of context

 • Scalability: support for multiple Telegram accounts and simultaneous chats

 • Isolated management of each chat — separate histories, styles, and contexts

• Secure storage of all data (correspondence, profiles, metadata) on the server

 • Ability to promptly set tasks and goals for the bot — for a day, scenario, or specific dialogue

 • Flexible management of communication style and behavior changes during the dialogue

 • Connecting an operator to the chat if necessary or upon trigger activation

 • Anti-spam logic with a "human" manner of communication: without templates, repetitions, with imitation of a live style

 • Trigger and alert system: keywords trigger automatic notifications in Telegram

 • Ability for manual prompting and setting communication goals directly during the dialogue

 • Support for proxies and IP anonymization for account security

 • Flexible architecture: scalability, connection of external logic, modularity

 • Integration with external knowledge bases, scripts, and client logic

 • Logging actions and analytics based on dialogue results

Additional wishes for the performer

Would be a plus:

 • Experience in creating Telegram userbots with support for multiple accounts

 • Skills in connecting and configuring OpenAI API or similar language models

 • Understanding of Telegram limitations and anti-spam mechanics (important to avoid blocks during active bot operation)

 • Ability to work with the logic of "human" behavior, variable responses, and adaptive scripts

 • Experience with proxies, anti-detect solutions, alert systems

 Budget

 • MVP: from $500 to $700 (expansion possible after successful MVP)

(depending on the proposed architecture, speed of implementation, and depth of memory/logic)

Collaboration format

 • Work through GitHub: providing a repository with documented code

 • Initial test launch on VPS or locally (to demonstrate functionality)

 • Feedback and technical support: from 7 to 14 days after release

 • If desired — phased implementation with subsequent scaling and introduction of new features