Budget: 168 EUR Deadline: 1 day
ready to help you out with this
have huge experience working in full stack
will send you previous work to make sure we match
3
0
Budget: 168 EUR Deadline: 10 days
Here’s how I would implement it:
I store the state in PostgreSQL, where each project operates as a Finite State Machine. This means that a task cannot skip a stage or get stuck in uncertainty. Deadlines are tracked through task queues based on Celery or Redis: when a task is created, an event is immediately scheduled for the deadline time. If the server restarts, the data will not disappear because the queue has a persistence parameter for permanent storage, and the bot will pick up the work from the same second.
When processing incoming data, the bot first determines the type of content and context, checking what stage the client is currently at and what action they expect in the waiting status. Validation occurs in two levels: first, the code checks the technical parameters of the file, and then the AI assistant confirms the content, for example, whether it is indeed the required document. If everything is fine, the status changes automatically; if not, the bot explains the error without closing the task.
I use AI for intent classification, data recognition, and answering typical questions. It is prohibited from making independent decisions about changing business logic or statuses. The AI only returns a structured JSON with data, which the code validates against strict rules before any action.
Working with deadlines is automated through events in the queue. As soon as the time is up, the bot messages the client after an hour, and then once every six to eight hours to avoid spam filters. If there is no response for more than twelve hours, the bot automatically sends a notification to the team's work chat. Any activity from the client instantly changes the status in the database, which automatically deletes all scheduled reminders for that task.
To avoid rewriting code when new processes arise, I will move all the logic of steps, document types, and message texts into a JSON configuration in the database. The code will only retain a universal engine that reads these schemas. Thus, a new checklist or change of deadlines is implemented simply by updating a record in the database without the need to deploy new code.
Regarding experience, I have created Telegram bots in Python for generating presentations, which used the Google AI Studio token and Pexels, and a bot for generating dialogues based on AI; I have also created Android applications.
1
1
Budget: 168 EUR Deadline: 1 day
Good day! I am ready to complete this project and have extensive experience in developing various applications.
The list does not show proposals concealed by the client or freelancer with a Plus profile, as well as proposals violating rules
