Engineer building developer tools, coding-agent integrations, and practical AI-assisted products.

I work across backend systems, developer experience, and practical AI integrations. Recent projects focus on helping developers understand codebases, monitor agent work and build LLM features that stay explicit, validated, and permission-aware.

• Code intelligence inside coding-agent workflows
• Agent/session infrastructure: notifications, long-running tasks, remote control and review loops
• Safe natural-language interfaces for Django and Django REST Framework applications
• Local-first product experiments around music libraries, publishing and personal workflows

• pi-fallow — TypeScript extension that brings Fallow codebase intelligence into the Pi coding agent through agent tools and slash commands.
• tasklight — Go CLI for running long developer tasks and getting notified when they finish, fail, or need attention.
• pi-tasklight — TypeScript extension that connects Pi coding-agent sessions to Tasklight desktop notifications.
• django-asklens — Python/Django package for safe natural-language querying over registered models using validated LLM query plans and read-only ORM execution.

• Bgera (bgera.app) — local-first macOS music library for audio analysis, similarity search and playlist building. Previously musicidx.
• Pi Deck — global session manager and remote companion for the Pi coding agent.
• retreat.day — agent-assisted publishing platform for thoughtful content about retreats, mindfulness and self-care.
• AskLens — product direction around natural-language data exploration for Django applications, currently published as django-asklens.

Some of these projects are in active development and will be public when they’re useful beyond my own experiments.

• Developer tools and code intelligence
• Coding-agent UX and infrastructure
• AI-assisted software without replacing engineering judgment
• Python/Django, TypeScript, Go, and Rust
• Local-first and privacy-conscious applications
• Learning in public through small, useful open-source projects

I have more than ten years of software engineering experience across backend systems, product development, infrastructure and technical leadership.

Recently I’ve been focused on making AI-assisted development more inspectable and dependable: giving agents better project context, keeping humans in the review loop and turning recurring workflow problems into reusable tools.

• Résumé: revazi.github.io/resume
• LinkedIn: linkedin.com/in/revazi