How Nebula's Customer Support Team Uses Agents — Live Workshop Recording

In this live workshop, Nebula staff engineer Eiman Abdelmoneim builds a team of AI agents from scratch to handle real-world business operations — all in under an hour. Follow along as we: — Design an end-to-end support pipeline (triage → knowledge base → proactive monitoring) — Connect agents to Gmail, Airtable, Supabase, and Stripe via one-click OAuth — Build a live landing page with table reservations for Cairo Kebab restaurant — Set up email support that reads and responds to customer inquiries — Handle real audience questions about scaling, skills management, and event-driven workflows Whether you're brand new to AI agents or already experimenting, this workshop shows you what "AI-first operations" actually looks like in practice. 📌 Try Nebula free for 30 days — email [email protected] with code from the session. CHAPTERS ======== 0:00 — Intro & audience warm-up 1:53 — Workshop goal: building an AI-first company 3:00 — Nebula workspace walkthrough 7:46 — The support pipeline architecture (triage → knowledge keeper → ops) 15:28 — Live demo: Cairo Kebab restaurant build kicks off 20:29 — Connecting tools via OAuth (Airtable, Gmail) 24:01 — Agent monitoring and job configuration 30:49 — Thread-based context and agent delegation 37:47 — Scaling with mini apps (Q&A) 40:32 — Event-driven triggers and webhooks 44:11 — Broad vs narrow agent instructions 46:35 — Managing skills and memory (skill creep, maintenance agents) 50:52 — Multi-client workspace strategy 56:26 — Webhooks explained (Safiya's question) 59:04 — Knowledge Keeper agent in depth 1:00:19 — Optimizer agents: autonomous improvement loops 1:01:35 — Trial offer and recording follow-up