jesse mahoney | Founder Machine Learning Engineer Stealth AI Infrastructure Startup
AI Startup
San Francisco, CAOn-siteFull-time$140k – $220k / year
About this role
Jesse Mahoney
Founder & Machine Learning Engineer | Agentic Systems, Authorization, Runtime Governance & Auditability, LLM Red Teaming | Building Real-World AI (stealth)
San Francisco Bay Area (US)
201 connections • 227 followers
About
At a Stealth AI Infrastructure Startup, my role as Founder and Machine Learning Engineer centers on creating advanced agentic AI systems. I focus on designing production-grade control layers for LLM-powered agent systems, implementing staged decision flows, and developing replayable evaluation frameworks to analyze agent behavior. My work emphasizes runtime governance, traceability, and failure clarity, with a commitment to long-term system reliability. My self-directed engineering education has allowed me to architect AI-driven decision platforms that integrate real-time data streams, predictive modeling, and automated risk controls in adversarial, high-stakes environments. I am driven by a passion for building explainable, reliable, and safe AI applications, with a focus on addressing real-world challenges through innovative system design and rigorous testing. Right now, a small group of people are watching me build this in public. We’re close to launch. If you’re interested in:• agent systems that don’t rely on vibes • governance that exists at runtime, not in PDFs • building infrastructure instead of slides. Follow along & if you’re top-tier talent who wants to work on something hard, foundational, and early. I'm always looking to connect — you’ll recognize why this matters.
Total Experience: 1 year and 5 months
Experience
Founder Machine Learning Engineer Stealth AI Infrastructure Startup - Stealth Startup (Current)
Nov 2025 - Present (7 months)
What I actually built:
- Designed and built production-grade control layers for LLM-powered agent systems
- Implemented staged decision flows to separate evaluation, authorization, execution, and review
- Built replayable evaluation frameworks to analyze agent behavi