Brandon Slicklein.
WorkScopesWritingInterestsAbout
Brandon Slicklein
Brandon Slicklein
New York City · AI-native systems
AuraPath AI · Co-FounderAnthropic Claude Community Leader · NYCBuilding toward an AI-only software factory

I build systems
that build.

Operator-turned-builder working at the edge of autonomous software. I design AI-native systems that do real, recurring work: production platforms, scheduled agents, and the loops that run them. Co-founder of AuraPath AI and Anthropic's Claude Community Leader for New York City.

See the work →About me
§01 · Selected work
01 · Active R&D

Autonomous Software Factory

AI-only software delivery: point it at a repo, set the objective, and an agent swarm plans, builds, and reviews the feature end to end.

TypeScriptAnthropic SDKMulti-agent
02 · Production

Political Intelligence Terminal

A Bloomberg-terminal for political giving, built so an environmental foundation can decide where its 2026-cycle dollars matter most.

Next.jsSupabaseClaude API
03 · Production

Conversational Analytics Platform

Ask a multi-venue hospitality group's data warehouse a question in plain English and get an answer, a chart, and a forecast back.

Next.jsFastAPIDuckDB
All work →
§02 · How I work

Discovery to production, then kept alive.

Five scopes of work, from finding the one workflow worth automating to keeping the shipped system healthy on retainer.

01 · AI-Native DiscoveryWe find the single workflow where AI changes the economics, then map exactly what it takes to ship it.
02 · Proof of ConceptProve the highest-value workflow on your real data, and get an honest go or no-go before committing to a build.
03 · Phased DeliveryProduction AI shipped in milestones, with Claude in the loop for the work that actually recurs.
04 · Enablement & WorkshopsHands-on sessions that get your team building with Claude Code and agents, not just watching demos.
05 · Managed RetainerWe keep the system healthy, on call, and improving, so the agents you shipped stay reliable.
§03 · Writing
Toward an AI-only software factory

Most teams are building AI-first. I am interested in the harder version: AI-only, where humans set the objective function and agents operate the entire loop.

June 2, 2026
Why I host Learn Claude Code in NYC

The fastest way to understand a tool is to teach it to a room of sharp people and watch exactly where they get stuck.

May 21, 2026
Building a Bloomberg terminal for political giving

What it takes to put Claude in the loop on a real decision: ranking races, scoring candidates, and recommending where a foundation's dollars matter most.

May 18, 2026
All writing →