JOHNKEARNEY.

AI Safety Engineer & Researcher

Building trust infrastructure for autonomous agents

Available for contracts
& consulting
Chicago, IL
Remote-first
Available for AI safety contracts, agent infrastructure consulting, and expert calls.
Remote-first. Chicago, IL.
What I do

AI Safety Consulting

Agent governance, policy engine design, risk classification, safety benchmarking, and audit trail architecture for teams deploying autonomous agents.

Contract / Retainer

Agent Infrastructure

Build MCP servers, policy gateways, zero-trust agent architectures, and cryptographic audit systems. From design through deployment.

Project-based

Red Team & Audit

Adversarial testing of AI agents and LLM systems. Prompt injection defense, tool-use safety evaluation, compliance gap analysis.

Engagement / Bounty

Expert Calls

Available through expert networks for deep dives on agent safety, MCP protocol, AI governance frameworks, and enterprise AI deployment strategy.

Hourly
What I've built
48
Autonomous bots
in production
2
Safety leaderboards
published
3+
Years building
production systems
250+
AI research papers
analyzed
2024 - Present

Authensor

Founder & Lead Engineer

Building the fail-safe policy gateway for AI agents. Every agent action gets classified by risk, routed through configurable policies, and stamped with a cryptographic receipt.

  • Designed and built risk-tier classification engine with configurable policy routing
  • Implemented cryptographic receipt system for tamper-proof audit trails
  • Built MCP-native integration (Authensor for OpenClaw) for agent policy enforcement
  • Architected zero-trust gateway: low-risk auto-passes, high-risk escalates to human, dangerous gets blocked
  • Developing A15 Certification program for enterprise AI safety compliance
PythonTypeScriptMCP ProtocolCryptographyPolicy EngineREST API
2024 - Present

15 Research Lab

Founder & Research Lead

Open-source AI safety benchmarking and research lab. Running transparent benchmarks, publishing open leaderboards, and providing the community with real safety data.

  • Built and maintain Agent Safety Leaderboard, ranking AI agent frameworks on safety metrics
  • Built and maintain MCP Safety Leaderboard, evaluating MCP server implementations
  • Indexed 250+ frontier AI research papers with automated analysis pipeline
  • Publish weekly frontier roundups, red-team roundups, and leaderboard updates
  • Developing A15 Certification for independent safety verification of enterprise AI vendors
PythonBenchmarkingRed TeamingResearchOpen Source
2024 - Present

Autonomous Bot Network

Architect & Builder

48-bot autonomous network running in production. Handles research indexing, content distribution, bounty hunting, code auditing, and cross-platform publishing, fully autonomous.

  • Designed modular 4-lane architecture: Research, Brand, Personal, Meta (orchestration)
  • Built bounty executor that clones repos, writes code, runs tests, and opens real PRs autonomously
  • Implemented cross-lane intelligence sharing via file-based brain system
  • Auto-distributes content to 9 platforms (Dev.to, Hashnode, LinkedIn, GitHub, etc.)
  • Production infrastructure: systemd service, auto-restart, health monitoring, cost tracking
TypeScriptNode.jsElectronLLM IntegrationSystems Architecture
2021 - 2024

Enterprise Systems & Operations

Systems Engineer

Built and maintained inventory management, ERP integrations, and operational infrastructure across multiple business domains.

  • Designed inventory tracking systems handling real-time stock across multiple warehouses
  • Built ERP integrations bridging legacy systems with modern APIs
  • Managed operational infrastructure, monitoring, and incident response
PythonSQLERPOperationsSystems Design
Technical skills
AI Security
Python
TypeScript
Policy Engineering
Agent Safety
MCP Protocol
Red Teaming
Cryptography
Benchmarking
API Architecture
LLM Integration
Network Security
Pen Testing
Vector DBs
Systems Design
Key projects

Authensor

Policy Gateway for AI Agents

Risk classification, policy routing, and cryptographic audit trails for autonomous agent governance. Zero-trust architecture with configurable action gating.

15 Research Lab

AI Safety Benchmarking

Open-source safety leaderboards for AI agents and MCP servers. 250+ indexed research papers, weekly frontier roundups, red-team reports.

Bot Network

48-Bot Autonomous System

Production autonomous network: research indexing, content distribution, bounty execution, code auditing. TypeScript monorepo with Electron desktop app.

Open Source Contributions

Community & Safety Tools

Active contributor to AI safety ecosystem. MCP server implementations, benchmark suites, and agent testing frameworks.

Let's work together

Available for AI safety contracts, consulting engagements, and expert network calls. Remote-first, based in Chicago.