I build practical AI systems, experimental tools, and product workflows at the intersection of language, structure, and software.

My work is focused on:

• AI-native interfaces
• LLM systems and constrained workflows
• code understanding and developer tooling
• turning vague ideas into usable products quickly

I explore how structured and semi-structured languages can serve as better control surfaces for models than ordinary natural language.

I work on prompting, memory, constrained generation, agent workflows, and interfaces that make model behavior more legible and usable.

I build small, practical systems for code generation, repo hygiene, scaffolding, patching, and workflow acceleration.

I use AI tools not just to generate output, but to compress the path from idea to working product.

• machine-readable interfaces for reasoning and retrieval
• experimental workflows for AI-assisted product building
• code understanding and architecture-aware tooling
• systems for safer, faster iterative development

• AGENTS.md generator — repo scaffolding and agent-facing workflow docs
• lab.abvx — ABV lab and small experimental surfaces
• ABV Creative — writing, research notes, and experiments

• Website: abvx.xyz
• Medium: medium.com/@abvcreative
• Substack: abvx.substack.com

• LinkedIn: linkedin.com/in/abvcreative