Education Plan Tracker

Overview

Backend service powering an apprenticeship‑tracking application used to record, evaluate, and visualise Handlungskompetenzen (practical competencies) over time. Built as a structured, API‑first service using Node.js/Express and designed for team‑scale collaboration and predictable delivery.

My Role

I served as the technical lead for the backend, shaping the project’s architecture, workflow, and engineering standards. My contributions included:

• Introducing and driving the adoption of Express.js for the backend
• Setting up a Docker‑based development and deployment workflow
• Establishing a monorepo structure for shared code and consistent versioning
• Leading the team into a Scrum workflow (two‑week sprints)
• Enforcing peer‑review culture, ensuring that nobody merged their own MR without review
• Providing architectural guidance during the early system design phase

Core Functionality

• Competency progression tracking: Students could navigate through their assigned competencies (e.g., A1 → A1.1), log completion, and attach evidence links.
• Role integration with internal systems: Worked with my coach to connect to the internal company server to determine whether a user was a coach or an apprentice.
• Progress storage via MongoDB: Stored entries, proofs, and competency updates; only minimal DB setup required beyond inserting and querying required data.

Features

• REST API for tracking competencies, learning goals, and progress entries
• MongoDB data model optimised for timeline‑based querying
• Role‑based access patterns for students, teachers, and administrators
• Structured evaluation logic to aggregate and represent competency development
• Modular controllers for clean separation of routes, logic, and validation
• DTO‑style request validation for predictable API behaviour

Architecture

The backend is structured into clear layers, making it easy to maintain and extend:

• Routing Layer: Express routers grouped by domain (competencies, users, tracking entries)
• Controller Layer: Encapsulates request handling and input sanitisation
• Service Layer: Contains the core evaluation logic and business rules
• Database Layer (MongoDB): Mongoose models with schema validation
• Shared Modules (Monorepo): Common utilities, typing, and config shared across services

Development Workflow

• Monorepo with shared modules for consistent tooling
• Dockerised environment for onboarding‑friendly setup
• Two‑week Scrum sprints, releasing to main after each sprint
• Mandatory code reviews to maintain quality across the team

Deployment & Operations

Although the company had no CI/CD pipelines, I ensured stable production deployments:

• SSH-based deployment workflow: Connected to the server, pulled updates, and coordinated releases.
• Docker Compose orchestration: Rebuilt images and launched the production stack with correct environment variables.
• NGINX reverse proxy + HTTPS: Configured NGINX for secure routing and TLS termination.
• Consistent release cycle: Deployed updates at the end of each two-week sprint.

Key Decisions

• Choosing Express.js for its simplicity, flexibility, and low learning curve for the team
• Adopting Docker early to eliminate environment drift
• Structuring the project as a monorepo to avoid duplicated logic and config
• Introducing Scrum and MR reviews to ensure steady, high‑quality delivery

Challenges / Lessons Learned

• Establishing a review culture required deliberate guidance, but dramatically improved quality
• Switching to Docker early reduced onboarding friction and environment inconsistencies
• The monorepo introduced a learning curve but simplified scaling as the project grew
• Early architectural decisions (Express + MongoDB) paid off by keeping complexity low