I'm a software engineer, independent researcher, and founder of Zhilo Labs.

I build AI agents, developer tools, and full-stack software. My interests are in agentic AI, long-horizon coding agents, memory systems, software reliability, and autonomous workflows.

I also conduct independent research on AI systems and the economic impact of automation.

You can reach me on LinkedIn or via email.

• REA: Reverse Engineer Anything - Concurrent configuration edit protection, canonical path handling, safe publication and conflict detection (#4855, merged). Bounded historical source imports, Git metadata limits, resource-constraint reporting, and regression tests (#7125, open).

• AI Job Search - Job scraping deduplication and context-size reduction (#710), robots.txt enforcement across HTTP redirects (#680), upstream fetch failure reporting (#681), and agent retrieval instruction corrections (#638). Also working on deadline-aware job ranking and retry fairness (#682, open), and isolated application-attempt tracking with atomic state updates (#781, open).

• Lexsus - Local-first AI coding bridge that connects web AI agents to local development environments through MCP. Preserves project context across agents, provides filesystem and terminal tools, and supports structured memory and secure handoffs.

• Zcron - AI-native gym operations SaaS for memberships, billing, attendance, staff management, analytics, and automated retention workflows. Built with Next.js, FastAPI, PostgreSQL, and React Native.

• Zhilo Labs - Software engineering and product development focused on AI-native applications, autonomous agents, developer infrastructure, and business automation.

My independent research focuses on long-horizon AI agents, memory reliability, software verification, and the effects of artificial intelligence on human work.

• Causally Sound Memory Invalidation for Long-Horizon Software Engineering Agents - Proposes CAVI, a framework for detecting invalid assumptions in persistent agent memory using causal dependencies, provenance, transitive invalidation, and uncertainty-aware verification.

• Can Humans Compete with Artificial Intelligence? - Research on AI-driven task automation, labor displacement, productivity, human-AI complementarity, and the future of work.

• Research Papers - Collection of my research manuscripts, preprints, and technical investigations.

• Reliable long-horizon software engineering agents
• Persistent memory, context engineering, and causal invalidation
• Multi-agent orchestration and tool-use reliability
• Agent evaluation, verification, and software correctness
• Local-first developer infrastructure and AI-native products