Backend Developer Data-Oriented

Our client, is a global leader in the development, integration, and implementation of advanced physical and cybersecurity, intelligence, and IT solutions, delivering complete end-to-end solutions on the enterprise and national levels. They are now looking for an experienced Backend Developer (Node.js, MongoDB, AI Agents) — Data-Oriented.

Location: Europe
Type: Remote, Full-time
Start date: ASAP

What you'll do (AI-first):

• Own and implement the platforms core AI capabilities, including: 
  • Lesson generation (structured lesson plans, sections, exercises, metadata)
  • Presentation generation (slide decks, visuals prompts, outline deck pipelines)
  • Content generation (explanations, examples, step-by-step guidance, assessments)
  • Auto-grading (grading logic, partial credit rules, rubrics, feedback generation)

• Build and maintain agent-based workflows to orchestrate multi-step AI tasks (tool calling, task pipelines, retries, evaluation).
• Design AI systems that are reliable in production: rate limits, fallbacks, model routing, prompt versioning, structured outputs, and validation/repair.
• Implement data pipelines that store AI outputs, revisions, and evaluation signals for continuous improvement.
• Build backend services in Node.js (JavaScript) and design scalable APIs to power AI features end-to-end.

• Work with unorganized / messy datasets and improve them over time (cleanup, normalization, migrations).
• Build analytics/BI-ready outputs from product and learning data (KPIs, segmentation, reporting endpoints).
• Optimize MongoDB performance (aggregations, indexes) and implement caching (Redis or similar) for hot paths.

Requirements:

• Strong backend experience with Node.js and modern JavaScript.
• Proven experience building AI-native product features in production (not just demos).
• Hands-on experience with agent frameworks / agentic patterns (multi-step orchestration, tool execution, workflow graphs, evaluation loops).
• Strong ability to implement structured AI pipelines: schema-driven generation, output validation, error recovery/repair, versioning, and observability.
• Experience with MongoDB (data modeling, aggregations, indexing, performance tuning).
• Experience handling messy/unstructured data and evolving schemas safely.
• Experience with caching systems (Redis or similar), including invalidation strategies and performance thinking.
• Strong reliability mindset: retries/timeouts, idempotency, background jobs/queues, monitoring/logging.

Preferred:

• Experience with grading systems (rubrics, partial credit, test-case style evaluation, calibration).
• Background workers/queues, streaming responses, event-driven architecture.
• CI/CD + automated testing for core workflows.
• Security best practices (auth, permissions, secrets management).

What success looks like

AI generation + grading flows are stable, fast, and consistent at scale.

Lessons/presentations/content pipelines produce high-quality structured output with strong guardrails.

The system gracefully handles model failures, rate limits, and edge cases.

Messy data becomes usable, and product insights are accessible for decision-making.