A web-based platform for automating the search and engagement of models in Instagram

General Description

The goal of the project is to create a web interface for automated search of models in Instagram, initial communication with them and collection of profiles. The system will use parsing by specified filters, messaging and AI agent for communication. All user (scout) actions are tracked and managed through a personal account with an analytics table (in the spirit of Airtable).

⸻

1. Users and authorization

Functions after logging in:

 - Access to your personal dashboard

 - Access to your own results table (similar to Airtable)

 - Ability to run through the control panel:

 1. Parsing

 2. Mailing

 3. autocommunication

 - Managing proxy and technical accounts

 - Tracking the status of processed contacts

 - Proxy management section (download, verification, validation)

⸻

2. Parsing by filters

Scout starts parsing through the interface by specifying the desired parameters. All filters can be enabled/disabled optionally.

Filters:

 1. Gender (male/female)

 2. Age (from - to)

 3. Content type (e.g., provocative / professional / lifestyle)

 4. Post texts (by keywords)

 5. Presence of links in biography (for links to aggregators and external platforms)

 6. Number of messages (from X)

 7. Posting frequency (date of last post)

 8. Number of followers

 9. Following certain influential persons (adult models, etc.)

 10. Geolocation

 11. Search by similar accounts (Look-a-like)

 12. Links to other platforms (TikTok, Twitter, etc.)

Result:

Collected profiles are automatically added to the table with values for each filter.

⸻

3. Processing table (similar to Airtable)

Each scout is provided access to their own interactive table:

 - Display of all found profiles

 - Fields with filters and the ability to evaluate the relevance of each

 - Columns with profile processing statuses:

 1) White: just performed parsing

      2) Yellow: mailing sent

 3) Orange: positive response received.

 4) Red: contact reached the questionnaire stage.

Additionally, you can implement color filtering, tags, comments, and sorting by activity.

⸻

4. Mailing

Mailing panel includes:

 - Field for uploading contacts from the table (for sending mailings)

 - Field for mailing message text

 - Ability to select the number of streams / Connect technical accounts

Additional feature:

Combined mode — launching parsing and instant mailing to "hot" contacts at their peak activity moment.

5. Automated communication

AI agent function, simulating live communication according to a given scenario:

 - The algorithm is built on the basis of dialogue examples (training on 50+ real conversations)

 - Goal: obtain consent to fill out a questionnaire (Google Form)

 - Optional: scout can use AI or communicate manually

 - Auto-replies are recorded in the table with updates
of status