System Designer / Python Developer / Parser Engineer

We are creating an innovative AI service for arbitration teams and media buyers that automates competitor analysis and creative generation.
The goal is to build a smart data pipeline that transforms raw creatives from spy services into unique structured records with metrics, performance ratings (0-100 points), and tags — for further analytics and model training.

In the future, the system will become the foundation of the AI Buyer — an LLM module that generates creatives based on keywords and optimizes campaigns in real-time.

Main Tasks

• Development of the data pipeline architecture for collecting, processing, and classifying creatives.

• Integration of spy services (via API or parsing) and normalization of data into a unified structure.

• Building a metrics system and rating evaluation of creatives.

• Development of tagging and semantic content analysis modules.

• Integration with LLM models for generating texts and visual elements (creative generation).

• Development of the basic architecture for further model training (fine-tuning, embeddings).
  
  

Technology Stack

Programming Language: Python 3.x

Back-end: FastAPI / Flask (for building REST API)

Data pipeline and processing: Pandas, Pydantic, asyncio, Celery (for asynchronous ETL processes)

Parsing / integrations:

Tyver API and other spy services (via API or web scraping)

Requirements

Experience in developing data collection, processing, and analysis systems.

Understanding of data pipeline architecture and ETL processes.

Experience in creating services or tools using LLM / AI models.

Ability to design scalable modular solutions.

Analytical thinking, initiative, readiness to build a system from scratch.

We offer

Flexible cooperation format (part-time / full-time / remote).

Opportunity to influence technical decisions and product architecture.

Payment — discussed individually