Development or configuration of parsing/grabbing Telegram chats

Data Parsing, AI Content Creation — incorrectly specified categories?

22 USD

Project translated automatically. Log in or register, to view the original

There are many open Telegram groups where people share interesting information. To save time searching for these curiosities (currently done manually), it is necessary to develop (or configure, if there is a ready-made solution) a parser/grabber that will download all the messages from the past day, and using LLM (Open Ai, Antropic, etc.) find something interesting in it. For example, someone mentioned that there will be an interesting event tomorrow, or that there is a recruitment for a children's gymnastics group. And if an interesting item is found, simply add it to a Google spreadsheet (or send it to a Telegram bot that will already add it to the spreadsheet via the SendPulse service).

I am looking for a performer who has relevant experience.

Here are some more details:

Project goal:
To create a system for automating the collection and sorting of useful information from Telegram groups. The main task is to reduce the labor costs of volunteers for monitoring information and provide journalists with ready access to important data for writing notes and news.

Task description:

Parsing Telegram groups:
- Approximate number — about 40-50 Telegram groups.
- It is important to automate the process of obtaining all messages from the last 24 hours for each group. All messages should be collected in one place for debugging and further processing.
Data processing:
- The collected messages need to be sorted into useful and unnecessary using a neural network, such as ChatGPT or another LLM (large language model). Information that we consider useful (can be added):
  - Real estate information (for example, messages about available apartments).
  - Medical services information (for example, doctor recommendations, information about temporary doctor absences).
  - Event announcements for children and adults (concerts, classes, clubs, etc.).
  - Ads for sale/giveaway/buy/search.
- The system must be able to automatically classify messages by categories (which can be added), for example: real estate, medicine, services, events, goods, advertising.
Information storage format:
- All useful information should be collected in a convenient form. Possible implementation options:
  - Google spreadsheet: Each message in a separate row with the following information:
    - Groups and cities from which the message was received.
    - Link to the message in the group.
    - Type of message (real estate, medicine, events, etc.).
  - Trello: Adding cards in Trello, where each card represents a message with the specified category and link to the original message.
  - Telegram bot: Sending filtered messages to the bot with the same information (city, message type, link to the original).
Implementation flexibility:
- You can propose your own options for data storage and processing based on your experience. The main thing is filtering information from a large number of groups and creating a convenient system for transmitting useful messages to journalists.
- Any suitable tool can be chosen for integration with LLM, for example, using OpenAI API to evaluate messages or another suitable solution.
- A semi-automatic solution is also possible if it reduces development time and costs.
Additional requirements:
- Regularity: The system must automatically parse chats every day and process data within a short time (up to a few minutes).
- Scalability: The ability to add new cities and groups in the future. Also, the ability to have more fine-tuned settings for message interests.

Thus, the main task is to create a system that will automatically parse messages in 40-50 Telegram groups daily, filter them through a neural network, and transmit useful information in a convenient format for the team.

Proposals 3

Please select

Mykyta Novytskyi

3 0

Projects 3
Rating -
Rating 548

Budget: 6999 UAH Deadline: 7 days

Hello!

I am ready to fulfill your request for creating automated sorting of information from Telegram chats. I have been creating Telegram bots for over 4 years.

Implementation plan:
• Create 5 Telegram accounts, which we will use to collect information from chats.
• Collect data from the chat for the last 24 hours with account rotation.
• Store the collected data in a database.
• Analyze messages in the database using chatGPT (sorting the messages you need).
• Save the information in a Google spreadsheet.

These are just the key points of the implementation.
Feel free to reach out, I will be happy to help.

Mikola S.

23 0

Budget: 7000 UAH Deadline: 5 days

Good day! I am ready to complete your project. I have experience working with the Telegram API, so I will do everything quickly and efficiently. Let's get in touch to discuss the details.

Mykhailo P.

123 0

Budget: 4000 UAH Deadline: 5 days

I have experience creating a parser that searches for job vacancies in chats and channels on TG. It uses openAI. However, the parser works in online mode (i.e., it checks messages as soon as they are published, not once every 24 hours).

Yakiv Tsypin
Kyiv, Ukraine

Projects 10
Contests 1
Rating -
Rating 372

Development or configuration of parsing/grabbing Telegram chats

Task description:

Mykyta Novytskyi

Mikola S.

Mykhailo P.

Proposals are currently absent

Current freelance projects in the category Data Parsing

Data parsing

Advertisement parser

Collection of contacts for newly established individual entrepreneurs

Creation of a stable parser/monitor for prices and availability for a demanding RTV/AGD store

Commodity nomenclature management system with flexible rights allocation