Skip to content

Vertical#

How Novakid runs 30 million live English lessons on OpenVidu

Novakid and OpenVidu customer success story Novakid and OpenVidu customer success story

Running a live video product where the users are seven years old is a challenge. Adults may tolerate a frozen frame or a reconnect spinner, but a seven-year-old just gives up, and the parent doesn't book a second lesson. That's what Novakid , an online English school for kids aged 4–12, has been doing since 2017. Over a million students, 50+ countries, more than 30 million lessons delivered, and up to 2,300 lessons running at once at peak.

We talked to Andrei Iakimov, DevOps Lead at Novakid, Inc., about moving their live classroom off a self-managed Kurento deployment and onto OpenVidu: why they migrated, why they didn't take the SaaS path, what that migration actually looked like, how they built scheduled autoscaling on top of it, what broke (and got fixed) along the way, and what they value most of.

How DynDevice Built Virtual Classrooms Into Its LMS with OpenVidu

DynDevice and OpenVidu customer success story DynDevice and OpenVidu customer success story

What do you do when your product depends on video meetings that happen somewhere else? For years, the trainers using DynDevice , the corporate eLearning platform built by Mega Italia Media , had to send their learners out of the LMS and into Zoom, Meet, Teams, WebEx or GoToWebinar links to run a live class. It worked — and it fragmented the learning experience every single day.

This post kicks off our series on how engineering teams solve real WebRTC and infrastructure challenges. We interviewed Matteo Resconi, IT & Development Area Manager at Mega Italia Media, about the journey from juggling five external meeting tools to one-click virtual classrooms built into their own platform: what they evaluated, why they didn't build on raw WebRTC, and what changed for their team and their users.

5 conversational AI app ideas you can build this weekend

Conversational AI app ideas for 2026 — real-time voice and vision agents you can build this weekend Conversational AI app ideas for 2026 — real-time voice and vision agents you can build this weekend

AI is moving faster than any of us can comfortably keep up with. Every week brings a new model, a new demo, or a viral thread claiming that everything has changed again. Keeping up feels almost impossible.

As AI becomes mainstream, the noise grows even louder. Social feeds are packed with promises of overnight success: build a startup with vibe coding, replace entire teams with a handful of prompts, make money while you sleep.

This post isn't another one of those promises.

Instead, you'll find five conversational AI app ideas you can realistically build over a weekend. Some are practical, some are a little unconventional, but all of them will teach you something valuable about how modern AI systems actually work.

More importantly, these aren't just toy projects. They can help you develop real AI skills, build a portfolio, test business ideas, and maybe even create something people are willing to pay for.

5 React video call platforms in 2026: Is SaaS still the right choice?

React video call platforms in 2026 — SaaS vs Self-hosted

1. Introduction

When React developers need to add video calls to their applications, the first question is usually simple:

"What is the fastest way to get it working?"

Most teams start by looking for a video API with a React SDK they can integrate quickly, without dealing directly with WebRTC complexity.

In practice, that usually means exploring well-known SaaS platforms that promise quick setup and minimal infrastructure work. That choice makes sense at the beginning:

  • Fast to integrate
  • No infrastructure headaches
  • Familiar turnkey experience

But there are questions teams often ask too late:

  • What happens when video becomes core to your product?
  • What happens when usage grows faster than expected?
  • Is paying per minute still the smartest choice in 2026?