Senior NLP and Python Developer

Department: e-Health & AI
Reporting to: Head of AI & Data
Working hours: Full-time, 9:00 AM - 6:00 PM CET

Project Overview
We are looking for a Senior NLP & Python Developer to join a product team building a Clinical Trial Management System (CTMS) for research sites.
The platform helps clinical research organizations activate and manage studies efficiently while ensuring full regulatory compliance. A major component of the system is an AI-powered protocol intelligence engine that analyzes complex clinical trial protocols (100–400+ page medical documents) and converts them into structured workflows.

The AI system extracts and processes:
* Schedule of Assessments (SoA)
* Visit structures and timelines
* Inclusion / Exclusion criteria
* Investigational product requirements
* Regulatory and operational tasks

This is a technically demanding environment requiring:
* RAG pipeline architecture
* medical document parsing
* structured data modeling
* deterministic AI output design
* traceability and auditability
* validation-ready engineering practices aligned with regulated environments

Tech Stack
Backend / Infrastructure
* Python 3.x
* FastAPI
* SQLModel / SQLAlchemy
* PostgreSQL
* pgvector
* Docker / Docker Compose
* Celery
* Redis
* RabbitMQ

AI / NLP
* LangChain
* LangGraph
* RAG pipelines
* embeddings
* vector search
* LLM orchestration
* NER models
* VLMs
* document extraction pipelines

Engineering / Quality
* CI/CD
* structured logging
* observability / monitoring
* automated testing
* secure coding practices

Team Structure
Cross-functional product team including:
* Senior ML/Python Engineer
* ML Architect
* Frontend Engineer
* QA Engineer
* Business Analyst
* Project Manager

Responsibilities
AI / NLP Architecture
* Design and optimize RAG-based pipelines for medical protocol parsing
* Refactor LangChain workflows for scalability, modularity, and cost efficiency
* Implement dynamic model orchestration based on task complexity
* Improve explainability and traceability of AI outputs
* Build deterministic transformation layers from extracted AI output to executable workflows
* Optimize vector search performance
* Integrate external regulatory and medical knowledge sources
* Develop document/content classification pipelines

Backend & Data Architecture
* Design scalable PostgreSQL schemas
* Model complex clinical entities:
  * Studies
  * Visits
  * SoA
  * Eligibility criteria
  * Regulatory tasks
* Ensure audit trail capabilities
* Improve database performance and transactional consistency

System Hardening & Validation
* Implement monitoring and structured logging for AI pipelines
* Improve reproducibility and explainability
* Expand automated testing coverage
* Support validation readiness for regulated environments
* Enforce secure engineering practices

DevOps & Scalability
* Maintain Docker-based deployments
* Improve CI/CD workflows
* Support monitoring and error tracking
* Ensure scalability for multi-site / multi-study usage

Required Experience
Backend / Architecture
* 5+ years with Python
* Strong FastAPI experience
* Deep SQLAlchemy expertise
* Strong PostgreSQL knowledge (schema design, indexing, migrations, tuning)
* Experience with Docker-based deployments
* Experience with Celery / Redis / RabbitMQ
* Strong system architecture skills
* Experience with document extraction systems

NLP / AI
* Hands-on experience with LangChain
* Production RAG implementation experience
* Experience with vector databases (preferably pgvector)
* Knowledge of embeddings and model orchestration
* Prompt engineering experience
* Experience designing deterministic AI systems

Domain Experience
Experience in regulated industries is strongly preferred:
* healthcare
* life sciences
* fintech
* other compliance-heavy environments

Nice to Have
* GxP / validation-ready AI systems experience
* experience with pre-commit / Ruff / Poetry
* large-scale system re-architecture experience

Ideal Profile
Looking for someone who:
* thinks architecturally, not just feature-by-feature
* can balance AI innovation with production reliability
* is comfortable in early-stage product environments
* communicates well across engineering, product, and QA teams
* has real production AI/NLP ownership, not only experimentation or prototypes



Start date: beginning of June
Duration: 6 months
Working hours: 9 AM – 6 PM CET
Payment terms: 40 days