RAG-PIPELINE
Public portfolio repo for Orion's operator stack — RAG indexing, LangGraph auto-fix, incident and watchdog modules. Published so reviewers can see how the agent is built and operated for Rightstune.
View repositoryAI engineering for music publishing — agents, pipelines, and production infrastructure.
"We are entering into the greatest transfer of wealth in history."— Maximilian Larson (Max1Millions) on AI
I'm currently specializing in agentic AI, context engineering, and agent harnessing. The global music publishing industry is one of the most fragmented data sectors in the world — so I started Rightstune, a publishing admin platform to help musicians get paid.
I also built Orion, an OpenClaw agent that gathers context from private repos using LangChain, modifies code locally via LangSmith and Claude, and calls production tools through MCP — ingesting CWR acknowledgements, processing royalty statements, and enriching catalog metadata autonomously.
Publishing administration, reimagined.
Rightstune was founded with a singular goal: help producers, artists, and musicians get paid their fair share of royalties.
Your music, your rights. Forever.
Production agent for Rightstune's Python, PHP, and SQL pipelines.
Orion is an OpenClaw agent I operate via iMessage. It maintains the full stack of publishing administration pipelines — from CWR generation to statement processing — with three interaction modes:
orion-rag-query for LLM-free Chroma + BM25 retrieval
orion-fix → LangGraph + Claude, human-in-the-loop branch review
Production runs on the Mac host via MCP; Orion's Linux sandbox never reads real .env secrets. These agent workflows run in production and have had measurable revenue impact for Rightstune.
orion-fixCollections: repos · docs · sql · playbooks · discrepancies
Public repos from github.com/max1millions.
Public portfolio repo for Orion's operator stack — RAG indexing, LangGraph auto-fix, incident and watchdog modules. Published so reviewers can see how the agent is built and operated for Rightstune.
View repositoryMaintainer of WESO's CISAC CWR Python domain model — configurable parsers for CWR files and model-JSON transformations used in global publishing registration.
View repositoryCLI validator for CISAC CWR .V21 and .V22 files — structural decode checks against the DataApi grammars, with production-file shims for common HDR edge cases.
Python CLI toolkit for the MUSO.ai v4 Data API — artist search, profile and track retrieval, credits listing, roles taxonomy, and daily quota tracking.
View repository