OI
All work
Backend2026

WriterIX AI Pipeline

Async AI microservice powering WriterIX content generation — FastAPI + Celery with Gemini/Groq LLM orchestration.

Role

Backend Engineer

Capacity

Internal service — WriterIX

Year

2026

Stack size

9 tools

Overview

The WriterIX AI Pipeline is a standalone FastAPI microservice that handles all AI-driven content generation for the WriterIX platform. It exposes HMAC-secured endpoints that trigger Celery task chains for blog post generation: topic research, outline creation, section writing, SEO metadata, and featured-image selection. LLM calls use Gemini as the primary model with Groq as a fallback, with structured Pydantic output parsing for reliable results. The service communicates generation events back to Laravel via webhook callbacks, with each pipeline step broadcasting a typed event so the frontend can render live progress.

Responsibilities

  • Designed the Celery task-chain architecture for multi-step blog generation
  • Integrated Gemini (primary) and Groq (fallback) LLMs with structured Pydantic output parsing
  • Implemented HMAC key verification and secure webhook callbacks to Laravel
  • Achieved a clean type-checker run with full lint compliance across all service files

Highlights

  • Multi-step LLM pipeline with Gemini/Groq failover
  • Typed event broadcasting for live progress in the dashboard
  • Comprehensive automated test coverage

Tech stack

PythonFastAPICeleryPydanticDockerRedisGemini APIGroq APIPytest