← All case studies
CASE STUDIES · 2026

PromptID

PromptID is an AI-native EdTech platform for employers and universities that examines learners for real, in-depth understanding, analysing the train of thought instead of rewarding memorisation. A proprietary algorithm drives the assessment engine under the hood.

AI 2026 Live
01 Overview

Overview

PromptID is an AI-native EdTech platform for employers and universities that examines learners for real, in-depth understanding, analysing the train of thought instead of rewarding memorisation. A proprietary algorithm drives the assessment engine under the hood.

02 The Challenge

The Challenge

Building an AI-native assessment platform from zero is complex enough. Add a hard requirement for +1,000 concurrent users, LLM model agnosticism, and genuinely intuitive UX throughout, and the scope gets serious fast. Everything had to be production-grade from day one.

03 What We Did

What We Did

We built the entire platform from scratch: NestJS backend, NextJS frontend, LangChain-powered LLM integration with model-agnostic architecture, Kubernetes infrastructure with up/down autoscaling, and a full observability stack, all benchmarked to +1,000 concurrent users.

04 Outcomes

Outcomes

6 weeks 0 to Production
LLM agnostic Via LangChain
Selected Screens
05 What We Learned

What We Learned

LLM model agnosticism pays off fast. Clients want flexibility as the model landscape shifts and LangChain's abstraction layer made swapping models a config change, not a rewrite. Kubernetes autoscaling also proved its worth: no overprovisioning, but the system scaled up under load in seconds.

Tech Stack
Tags
AIEdTechLLM

Want outcomes like this?

Tell us what you're building. We'll tell you whether we're the right team for it.

 Book a call