Esoteric Labs
Esoteric Labs is an independent AI studio that deconstructs complex, specialized and regulated domains and builds infrastructure and AI systems rather than only applications. Its stated approach is to build from deep understanding, decomposition and structured experimentation.
At a Glance
- Organizations with specialized, regulated or otherwise high-complexity workflows
- Software teams and developers using Claude Code, Codex or other coding-agent runtimes
- School psychologists and special-education assessment workflows
- Healthcare, legal, financial and insurance documentation teams
- +2 more
AI Tools by Esoteric Labs
(1)
OpenRig
Open Source Multi Agent Harness
Discussions
No discussions yet
Be the first to start a discussion about Esoteric Labs
Latest News
OpenRig v0.6.1 released with first-use repairs, improved Codex readiness, linked-worktree launches and a consolidated help path.
OpenRig v0.6.0 released as a stranger-ready npm release with per-seat permission choices, typing guard, Herdr launch action and runtime improvements.
OpenRig v0.5.17 added Bun installation support and removed the unpublished daemon package dependency.
OpenRig v0.5.16 added selection of CLAUDE.local.md for managed Claude instructions and an advisory portability report for pull requests.
Products & Services
Open-source Apache 2.0 multi-agent harness/local control plane for coding teams. It defines agent and rig topologies in YAML and provides a daemon, CLI, MCP server and React/terminal interfaces for launching, coordinating, communicating with, discovering, snapshotting and restoring Claude Code and Codex sessions; Pi and OpenHands support are in development.
Personal coordination system for AI-driven software development: a TypeScript CLI, Claude skills and filesystem-based workflows for knowledge compression, agent handoffs, scope management, task tracking and inter-agent messaging.
Production AI system for school psychologists that synthesizes assessment documents and generates legally binding disability-qualification reports while handling district-specific requirements and many output configurations.
Food-science agent with self-improving evaluation loops for recipe optimization.
Market Position
Esoteric Labs positions its work around infrastructure and reusable primitives for difficult domains rather than generic applications. OpenRig is framed as an open-source, local control plane for developer-run multi-agent topologies that can combine Claude Code and Codex, contrasting with single-harness/sub-agent approaches and with cloud/proprietary options such as Anthropic Managed Agents; the studio also names Paperclip, Superset and Emdash as other multi-agent architectures to compare against.
Leadership
Founders
Mike Schwarz
Applied AI Product Engineer and the site's sole named operator. His background includes early data-center, Windows NT and network-engineering work during the dot-com era; he says he has spent roughly 8,000 hours coding with AI and previously built the private Agent Focus harness. GitHub identifies him as an Applied AI Product Engineer at Esoteric Labs.
Executive Team
Mike Schwarz
Applied AI Product Engineer; Founder and creator
Leads Esoteric Labs and its product work. He describes extensive AI-assisted production coding, created the private Agent Focus framework and is the creator of OpenRig.
Founding Story
The studio grew from Mike Schwarz's practice of demystifying difficult infrastructure and AI problems and from building production software with AI agents. OpenRig emerged after he spent about 8,000 hours coding with AI and needed a durable way to snapshot, coordinate and recover networks of specialized coding agents; Esoteric Labs describes its broader vision as building infrastructure whose quality compounds and makes the next hard problem easier.
Business Model
Revenue Model
Independent consulting and product development; the official site says Mike Schwarz is available for consulting and product development. OpenRig is distributed as open-source Apache 2.0 software via GitHub and npm.
Target Markets
- Organizations with specialized, regulated or otherwise high-complexity workflows
- Software teams and developers using Claude Code, Codex or other coding-agent runtimes
- School psychologists and special-education assessment workflows
- Healthcare, legal, financial and insurance documentation teams
- Food science and recipe-development teams
- Environmental monitoring and climate-analytics teams
- Multi-week or multi-month software development by coordinated coding-agent teams
- Software factories and persistent local agent fleets
- Complex regulated documentation in psychology, healthcare, legal, financial and insurance workflows
- School-psychology disability qualification reporting
- Recipe optimization and food-science experimentation
- Real-time environmental and climate monitoring