Hack & Tell: Multi-Model Architectures & Voice AI Observability

Hey peeps!
Join us for an evening Hack & Tell with Serhat Soydan and Arnaldo Vera at Harbour.Space Institute of Technology in Poblenou - two focused, practical talks plus plenty of time to meet other AI engineers and founders!

Serhat will share how he built and scaled a multi-model AI product as a solo founder, while Arnaldo will show how Enera’s Voice AI stack uses observability and latency metrics to keep real-world AI systems fast and reliable.

🗓 Agenda – Tuesday, December 9
18:30 – Doors open, drinks & arrivals
19:00 – Talk 1: Serhat Soydan, Talk 2: Arnaldo Vera
20:30 – Open networking

📍 Location
Harbour.Space Institute of Technology
Carrer de Rosa Sensat, 9-11, 08005 Barcelona
6th floor
🎟 Cost: Free (RSVP required)

***

🎤 Speaker 1: Serhat Soydan - Building GenAIntel: multi-model AI in production as a solo founder
Serhat will show how he built and scaled GenAIntel.com as a solo founder, from Next.js + Azure + Fal.ai to supporting 100+ models in production. He’ll share the key architecture choices, what actually worked with coding models like Cursor + Claude/Gemini, and a few things he’d do differently next time.

🎤 Speaker 2: Arnaldo Vera - What Voice AI can teach us about observability
Arnaldo will unpack Enera’s Voice AI stack and the metrics that really move latency: context windows, memory/cache, network hops, and model choices. He’ll demo the observability setup they use in production and the practical “knobs to tweak” that transfer to any serious AI system.

See you soon!