Natnael Alemseged
AboutProjectsTestimonialsWork Experience
© 2026 Natnael Alemseged. All Rights Reserved.
Secure Agent Protocol // Latency Critical // Addis Ababa

TRP1 AI Artist

Generative Media Infrastructure Engineer

Async multi-provider AI content generation framework for music, video, and images with plugin providers, style presets, CLI workflows, job tracking, duplicate detection, and cost controls.

"Unified music, image, and video generation into a provider-agnostic pipeline with persistent job management."
TRP1 AI Artist multi-provider content generation pipeline
Click to Zoom
Async music, image, and video generation with provider plugins, job tracking, cost controls, and media export

Problem

Generative media APIs have inconsistent capabilities, long-running jobs, duplicate-cost risks, provider lock-in, and messy post-processing paths from source material to final output.

Solution

Built a typed plugin architecture for Google, AIMLAPI, and Kling providers, plus style presets, Archive.org source discovery, SQLite job tracking, async status sync, FFmpeg media merge, and export workflows.

Deep Dive

What It Generates

TRP1 AI Artist orchestrates music, image, and video generation across multiple providers. It can search source material, run parallel generation jobs, merge media locally, and prepare outputs for YouTube or S3.

Engineering Highlights

  • •Provider registry: new music or video providers can be added without modifying core orchestration.
  • •Job tracking: SQLite persistence records queued, processing, completed, downloaded, and failed jobs.
  • •Cost control: duplicate detection prevents repeated API calls for the same prompt and provider setup.

Tech Stack

PythonTyperPydanticSQLiteFFmpegAsyncIO

Tags

#Generative Media#AI Art#Provider Plugins#Creative Tools
View GitHub Repo

More CreativeWork Software

Case studies in similar engineering domains.

DataAgentBench Evaluation Fork

→

Fork and evaluation workspace for DAB, a realistic enterprise data-agent benchmark spanning multi-database integration, messy joins, unstructured text transformation, and domain knowledge.

SalesConversion-Bench

→

Sales-domain benchmark and trained critic layer for Tenacious-style B2B outreach, with contamination-aware task generation, deterministic scoring checks, preference data, and a small LoRA judge.

Project Chimera

→

Spec-driven autonomous influencer network foundation using FastRender Swarm architecture, MCP-only external IO, Planner/Worker/Judge task DAGs, HITL review, multi-tenancy, and budget governance.