Full Stack Software Engineer - AI

This role is ideal for a Fullstack Software Engineer with 6+ years of experience, a strong foundation in AI/ML, and hands-on skills in React, Node.js, Python, and cloud platforms like AWS. Best suited for someone passionate about building generative AI applications and conversational tools, with knowledge of LLMs, RAG, and scalable system design. It's a great opportunity for engineers based in the Netherlands who want to innovate in the learning and development space.

1. How would you architect a fullstack AI product like CAISY using React, Node.js, and LLMs?
   Use React for a dynamic conversational UI, Node.js for API orchestration, integrate LLMs (like GPT via OpenAI/Azure), store conversation history in PostgreSQL or a vector DB, and implement async tasks with RabbitMQ or SQS.

2. Describe your experience integrating LLMs into a product for real-time user interaction.
   [Example] Used OpenAI's GPT API with prompt engineering, managed user sessions and rate limits, handled fallbacks and token errors, and provided real-time feedback with React frontend updates.

3. What’s your approach to ensuring secure and private communication in an AI-powered SaaS tool?
   Apply token-based auth, encrypt sensitive data in transit and at rest, sanitize inputs to LLMs, redact PII, and comply with GDPR or similar data privacy laws.

4. How do you build scalable microservices for AI workloads?
   Use containerized Node.js services in Kubernetes, design stateless APIs, offload AI inference to background tasks when needed, and use Kafka/SQS for task orchestration.

5. How do you ensure real-time conversational performance using generative AI and React?
   Stream LLM responses (e.g., OpenAI's streaming API), debounce UI inputs, show typing indicators, and cache frequent queries with Redis or a vector index.

6. Describe your experience with vector databases and retrieval-augmented generation (RAG).
   [Example] Embedded user content into vectors with OpenAI or Sentence Transformers, stored in Pinecone/FAISS, queried via semantic similarity, and passed results to GPT context window.

7. What’s your strategy for developing and testing fullstack AI features in a collaborative team?
   Work in Agile sprints, write contract tests for APIs, use mocks for LLMs during frontend testing, run CI with Docker, and pair code reviews with design walkthroughs.

8. How do you manage data pipelines for AI/ML in a Node.js environment?
   Use ETL scripts with Python or Node.js, schedule jobs via AWS Lambda or Airflow, store outputs in S3/Postgres, and trigger retraining or index updates as needed.

9. What are the challenges of vertical vs horizontal scaling in microservices, and how do you handle them?
   Vertical scaling hits resource ceilings; prefer horizontal scaling via Kubernetes replicas, ensure stateless services, and scale based on CPU/memory or queue depth metrics.

10. How do you explore and integrate new generative AI tools or libraries into your development workflow?
    Evaluate APIs with quick POCs, assess latency/cost, test reliability with fallback logic, document findings, and demo internally to drive cross-team adoption.

Skillsoft is hiring a Full Stack Software Engineer in the Netherlands to build AI-powered learning tools like CAISY—an innovative generative AI simulator for practicing workplace conversations. You’ll work with React, Node.js, Python, and AWS to develop scalable features using LLMs, vector databases, and modern data pipelines. Ideal candidates have 6+ years of experience, familiarity with AI/ML, and a passion for transforming learning. Join a global leader in edtech and shape the future of AI-driven professional development.