AI-powered learning platform
Scaling an AI-Powered Learning Platform to 8M+ Users
AI-powered learning platform. Founding engineer engagement plus VertCode contracting team—shipping AI/LLM features and scaling infrastructure from thousands of users to 8M+.
Founding engineer (Wesley Breukers) + VertCode contracting team
Jan 2024 – Jul 2026
AI/LLM features, full-stack engineering, scaling infrastructure
Die Herausforderung
StudyFetch set out to turn any course material into interactive, AI-powered study tools. The hard part was never the demo—it was making those AI features reliable and fast for a user base that grew from the low thousands into the hundreds of thousands, and eventually past 8 million.
Growth like that breaks things: database load spikes, latency creeps in, and features that worked fine for a few thousand users start to strain. The platform needed engineers who could ship new features quickly and keep the whole thing standing while it scaled.
What we did
Wesley joined as a founding engineer and worked across the entire stack, whatever it took to get a feature live and keep the platform healthy. That meant moving fast from planning to building to firefighting: shipping new product and AI features, then diagnosing and fixing issues like high database spikes under load before they became outages.
As the work grew, VertCode brought on a senior contracting team to build alongside the platform's needs. The work spanned:
- Building and shipping new product and AI features end to end
- Diagnosing and resolving performance bottlenecks, including high database load spikes, to keep the platform stable as traffic climbed
- Fast iteration across planning, implementation, and fixes, so features reached users quickly
Die Ergebnisse
Over two and a half years of sustained engineering:
- Scaled from thousands of users to 8M+
- Shipped AI study features that measurably improved learning outcomes: students who used the AI as a thought partner were 2.3x more likely to answer correctly
- Kept a rapidly growing platform stable through the scaling curve, shipping new features the whole way
Why it mattered
This is the kind of work VertCode does best: not slideware or proofs-of-concept, but production engineering that holds up under real scale and real pressure. Taking AI features from idea to millions of users, and keeping them fast and reliable while the platform grows, is exactly what we bring to the companies we work with.