Experience

AEOS LABS minimalist white logo on a black background

AEOS Labs — Full Stack AI Engineer

Sept 2025 – Present

// role: Full Stack AI Engineer

// domain: B2B Procurement + RFQ Automation

// stack: TypeScript · Next.js · Node.js · Hono · BullMQ · Prisma · PostgreSQL · Redis · Google Gemini · OpenAI · Microsoft Business Central · Microsoft Graph · Azure Container Apps · Azure Blob Storage · Azure Communication Services · Docker · GitHub Actions

  • A production B2B procurement platform that automates the full RFQ lifecycle, from a buyer submitting a parts request to vendors receiving quote requests, submitting quotes, and purchase orders being placed and tracked.
  • A Next.js frontend (Shadcn/ui, Redux Toolkit, TanStack Query) covering marketplace browsing, RFQ submission, vendor management, and quote review workflows, built from UX designs to production.
  • An AI-powered quote extraction pipeline that parses inbound vendor emails and document attachments automatically, cutting quote turnaround from 48+ hours to minutes.
  • A vendor sourcing orchestrator that cascades through catalog matching, preferred vendor lookup, and AI-driven web discovery before escalating to manual review, so procurement keeps moving even with incomplete data.
  • A bidirectional ERP integration with Microsoft Business Central, syncing vendors, items, quotes, and conversation threads in both directions, with Microsoft Graph webhooks for real-time inbound email ingestion.
  • Separated the platform into two independently deployable services: an HTTP API for user-facing traffic and a background worker for all async processing, so neither blocks nor scales with the other.
  • Treated Business Central as the financial system of record and the platform as the workflow layer, keeping concerns cleanly separated while maintaining consistency through bidirectional sync.
  • Added circuit breakers on every external service call and dead-letter queues on every job, so a third-party outage degrades gracefully instead of stalling the whole platform.
  • Deployed on Azure Container Apps with per-service scaling policies, Azure PostgreSQL, Redis, and Blob Storage, giving each layer independent scaling without shared infrastructure constraints.