Ableton MCP Server
Drive Ableton Live from natural language — transport, tracks, and devices exposed to Claude over MCP.
Python · MCP · Ableton Live API// 2025
Brent Piñero
> AI systems engineer · BAP Labs
Building multi-agent AI systems from real problems, not trending frameworks.
REC 00:14RECagent_trace
$ cat ~/career.log▍
2025 – Present
2023 – 2025
2022 – 2023
2018 – 2022
2015 – 2018
2013 – 2014
$ ls ~/recent_work▍
$ ls ~/builds▍
REC 00:08RECretrieve.py
Blender 3D Design MCP
Model and script 3D scenes by prompt — an MCP bridge into Blender’s Python API.
Python · MCP · Blender bpy// 2025
C.O.R.E — Client Outcome Review Engine
Turns raw client signals into structured outcome reviews — an LLM pipeline for engagement health.
Python · LLM · RAG// 2025
Lightroom Style Transfer
Learn a look from a reference frame and apply it across a catalog — style transfer for photo edits.
Python · Computer Vision// 2024
Winloop Task App
A task app built around win-loops — momentum from small, completed cycles.
TypeScript · React// 2024
HVAC Field Service App
Field-service app for HVAC technicians — scheduling, work orders, and on-site capture. Shipped as an iOS app.
Swift · iOS// 2023
$ whoami▍
My career started in finance and competitive intelligence — analyzing markets, building models, advising C-suite executives. But I kept noticing something: the most valuable insights came from systems I built myself, not tools someone sold me.
That pattern defined my path. At every company — C-III, Dallas Jet, Appian, PwC, Adobe — I ended up building the AI systems that made the intelligence work actually scale.
At PwC, I built a multi-agent system that won first place in their GenAI hackathon. At Appian, I shipped their first internal RAG chatbot before RAG was even a standard pattern. The through-line is always the same: spot what's emerging in AI before it goes mainstream, then build something real from it.
When I'm not building AI systems, I'm producing electronic music in Ableton Live — bass music, EDM, the kind of sounds that hit different at 2 AM. It all comes from the same place: AI engineering, audio tools, and the belief that the best software is built by people who actually use it.
— Brent