AI-Powered Employee Engagement & Gamification Platform
An HR technology company needed a platform that uses AI to drive employee engagement through personalized daily challenges, intelligent quizzes, real-time coding assessments, and gamified rewards — boosting workplace productivity and satisfaction.
プロジェクトを相談する
課題
Traditional employee engagement tools lacked intelligence and personalization:
- Generic Content — One-size-fits-all challenges and quizzes failed to engage employees with varying skill levels and roles
- No Adaptive Learning — Assessment systems couldn't adjust difficulty or content based on employee performance history
- Limited Assessment Types — Existing platforms supported only multiple-choice quizzes, not coding challenges or scenario-based questions
- Low Participation — Without gamification and social features, engagement programs had poor adoption rates
- Scalability — The platform needed to serve organizations of different sizes with isolated, customizable experiences
私たちのソリューション
We built an AI-powered engagement platform using OpenAI and Amazon Bedrock for intelligent content generation, Judge0 for sandboxed code execution, ChromaDB for RAG-powered content retrieval, and a full gamification layer with social features.
Architecture
- Backend: NestJS with microservices architecture
- Frontend: React with responsive design
- AI Models: OpenAI API and Amazon Bedrock (Llama) for content generation, quiz creation, and answer evaluation
- Code Execution: Judge0 API as sandboxed engine supporting 60+ programming languages
- Vector Database: ChromaDB for semantic search and RAG workflows
- Database: MySQL for relational data
- Container Orchestration: Amazon ECS with Docker
- Frontend Hosting: AWS Amplify
- CI/CD: GitHub Actions with automated pipelines
- Authentication: OAuth integration for enterprise SSO
Key Features
- AI Quiz Generation — Automatically creates quizzes tailored to each employee's role, skill level, and learning goals using OpenAI and Amazon Bedrock
- Adaptive Assessment — Analyzes past quiz results and engagement patterns to generate quizzes targeting weak areas
- Online Coding Platform — Embedded IDE with Judge0 API supporting 60+ languages, syntax highlighting, test case validation, and live output
- Coding Assessments — Timed coding challenges with automated scoring and detailed performance analytics
- RAG-Powered Content — ChromaDB stores embeddings of training materials and past assessments for intelligent content retrieval and recommendation
- Social Feed — Facebook-style feed where employees share achievements, post updates, comment with threaded replies, and react to posts
- Leaderboard System — Real-time rankings by points, challenge completions, and quiz scores across individual, team, department, and organization views
- Gamification — Points, achievement badges, team challenges, competitions, and progress tracking
- AI Answer Evaluation — LLM-based assessment of open-ended answers with contextual grading rubrics
- Admin Panel — Creating, managing, and assigning assessments to employees and teams with analytics dashboards
成果
技術スタック
caseStudyDetail.more ケーススタディ
その他の技術実装事例をご覧ください
AIを活用したOCRによる請求書処理とQuickBooks連携
毎月数百件の仕入先請求書を処理する中規模企業が、AI/OCRを使用して請求書データを自動抽出し、それを記帳と支払追跡のためにQuickBooksに直接同期させることで、手動データ入力を排除する必要がありました。
SCTE-35マーカー解析とマルチプラットフォームプレイヤー統合によるクライアントサイド広告挿入 (CSAI)
あるビデオストリーミングプラットフォームは、ウェブ、モバイル、コネクテッドTVアプリ全体でクライアントサイド広告挿入 (CSAI) を実装する必要がありました。これにより、サーバーサイド挿入では提供できない、完全な広告インタラクションサポート（クリック可能なオーバーレイ、コンパニオンバナー、スキップボタン）を備えた、パーソナライズされたデバイスレベルの広告体験が可能になります。