AI Qadam Meetup #3: what stands around the model

When we talk about AI, we almost always talk about the model. But launch a real product and you find the model is only its tip: beneath it sit the data you have to understand, the context without which answers meananothing, the people who have to use it, and the hardware that carries it all. On October 17 in Tashkent, at the third AI Qadam meetup, four speakers will show these layers from the inside, on live projects. AI Qadam is a community of AI practitioners in Central Asia.

🎙 Mokhiyakhon Uzokova, Linguistic Analyst, U-BSS
"Voice AI is not just a model"
Why "90% accuracy" in speech recognition guarantees nothing. Using Uzbek ("kerak edi" or "kerakdi"), we'll look at how speech data is built and why it decides whether a voice AI will sound human.

📊 Anna Shibanova, Partner at A+ Solutions, Data Community Uz
"Is your data ready for AI?"
Collecting data isn't enough: AI needs context. Anna will walk through the path from AI-Ready Data to Contextual Intelligence, using examples from corporate data and the financial sector.

🏗 Bakhtiyor Bekmurodov, 3D Visual Architect
"Architecture × AI: a new approach to architectural design"
First the tool: the path from AS IS to TO BE and beyond, to futuristic scenarios. Then the key part: how AI changes the way developers and clients communicate, a new format for our country.

🖥 Evgeniy Vishnevskiy, CEO of Pro-Data
"How to save money on AI infrastructure"
How to choose a GPU for local models, and how not to overpay for everything that has to surround it: the infrastructure without which the hardware doesn't work.

📅 Saturday, October 17, 1:00 PM
📍 Tashkent, Beeline Uzbekistan office, 4/4 Muminova St. (near IT Park)
👉 Registration in Telegram bot