Portfolio
Back to Projects

Aug 2025 – Oct 2025

MerAI

AI Career & Learning Mentor Platform

Next.jsReactPostgreSQLGoogle Gemini AIVercel

Overview

A full-stack AI mentoring platform delivering adaptive mock interviews and personalised skill assessments — cutting learner prep time by 45% in a 15-user pilot.

Problem Statement

Most career prep tools are static and one-size-fits-all. Learners need adaptive feedback that adjusts to their specific skill gaps and target roles.

Key Highlights

Adaptive mock interviews and skill assessments powered by Google Gemini AI, cutting learner prep time by 45% in a 15-user pilot

PostgreSQL backend with role-based dashboards and progress tracking, sustaining query response times under 200ms under concurrent load

Lighthouse score above 92 with edge-optimized routing and zero critical failures across production sessions

Architecture

Next.js App Router with PostgreSQL for user state and progress. Gemini AI handles interview question generation and answer evaluation in a structured prompt pipeline.

Challenges

Designing adaptive interview logic that feels dynamic rather than templated, and managing PostgreSQL query performance under concurrent load during the pilot.

Learnings

Conversational AI system design, database performance optimization, and the nuances of building AI-driven personalization that genuinely adapts to the user.