At work
Building Prosperna Commerce Copilot, a production AI system for merchant workflows.
Briuwu / Applied AI Engineer
Brian Laurence E. Millonte
I'm Brian. I design and build AI products people actually use.
Hi, I'm Brian, an Applied AI Engineer focused on building production-grade AI products that solve real business problems. I design and ship agentic LLM systems end-to-end, from backend orchestration and guardrails to streaming UX and reliable deployment.
Personal build
Designing AIRI, a local-first Windows assistant with approvals, traces, and typed tool flows.
Focus
Multi-agent systems, streaming interfaces, and human-in-the-loop product design.
Current work
Applied AI Engineer at Prosperna
Building production AI workflows for content, analytics, approvals, and multi-agent commerce tooling.
Flagships
- AIRIWindows-first local desktop assistant
- Prosperna Commerce CopilotMerchant-facing production AI system
Selected
Work
AIRI
A local-first Windows assistant built around an overlay UI, guided runtime setup, and approval-aware tool use instead of a generic chat shell.
- Built dual overlay modes with hotkey-driven focus handoff and chip-on-blur behavior for low-friction invocation.
- Implemented phase-based onboarding that checks runtime readiness before handing users into the overlay.
- Separated streamed reasoning from final content and paused tool execution for approval on safety-sensitive actions.
Prosperna Commerce Copilot
A production AI copilot that routes merchants between a general orchestrator and specialist agents for insights, merchandising, copywriting, and page building.
- Worked on a unified copilot that connects branding and media tools with specialists for analytics, merchandising, copywriting, and page building.
- Implemented streaming chat behaviors that keep widgets, approvals, pickers, and longer-running actions coherent in the UI.
- Built guardrails around branding readiness, plan access, active-agent ownership, and publish-safe workflows.
Current
Experience
June 2025 - Present
Applied AI Engineer at Prosperna
Built and scaled a production AI copilot for a commerce SaaS platform, enabling merchants to generate and refine product content, SEO metadata, blog content, landing pages, media updates, and business insights through a unified conversational experience.
- Architected a centralized multi-agent backend in Python and FastAPI with agent routing, tool orchestration, prompt versioning, and structured runtime contracts.
- Implemented real-time LLM streaming with Server-Sent Events, including chat state sync, widget flows, approvals, and artifact updates.
- Designed human-in-the-loop guardrails for high-impact operations through approval gates, guided selections, memory confirmation, and draft or publish checkpoints.
- Built session recovery, locking, fallback handling, and usage safeguards so longer AI workflows remained dependable in production.
- Integrated live business analytics so users could query traffic, sales, leads, products, and content performance in natural language.
- Worked across React frontend integration and automated testing for routing, streaming, restore, approval flows, and edge cases.
Capabilities
Stack
AI systems
Frontend systems
Backend and infra
Now
Focus
I work from Cavite, Philippines on AI products that need to do more than look impressive in a demo. Most of the work I care about now lives at the intersection of orchestration, interaction design, and production reliability.
- Shipping AI copilots that combine specialist agents, structured workflows, and clear approval moments.
- Designing streaming interfaces that keep long-running model work readable instead of opaque.
- Building local-first assistants that stay useful because state, logs, and recovery paths are explicit.