Senior NLP Engineer — Technical Knowledge Systems

Mia Johansson

Full-time · Senior · Paris

This role is no longer accepting applications

Similar open roles you can apply to now:

About the role

We are the R&D division of an aerospace manufacturer. We employ 4,800 engineers across six countries. Our engineers produce and consume an enormous volume of technical documentation — design specifications, maintenance procedures, failure mode analyses, airworthiness directives, and supplier qualification reports. Finding the right document, the right version, and the right passage within it takes a disproportionate amount of engineering time. We have been working for two years on a technical knowledge retrieval system that would let our engineers query across our full documentation corpus in natural language and receive accurate, cited answers. We have a working prototype. It is not good enough. Retrieval precision on highly technical aerospace terminology is lower than we need it to be, and the system does not handle multi-document synthesis questions — the kind where the answer requires combining information from a structural specification, a material datasheet, and a maintenance procedure. We are looking for a Senior NLP Engineer to own the technical architecture of this system and lead the work to make it good enough for production deployment in a safety-critical engineering environment.

Responsibilities

Own the technical architecture of the knowledge retrieval system — retrieval pipeline, generation layer, and evaluation infrastructure
Improve retrieval precision on aerospace-specific terminology through domain-adapted dense retrieval models
Extend system capability to multi-document synthesis questions — designing retrieval strategies, context window management, and faithfulness verification
Build evaluation frameworks appropriate to safety-critical deployment: attribution accuracy, hallucination detection, and citation quality
Work with aerospace domain experts to annotate evaluation datasets and validate system outputs against engineering ground truth

Requirements

6+ years of NLP engineering with production experience in technical document retrieval or question answering over specialised corpora
Python at a senior engineering level — production-quality, auditable code appropriate to safety-critical engineering environments
Hugging Face Transformers for fine-tuning dense retrieval and reading comprehension models on domain-specific technical text
RAG pipeline design at depth: hybrid retrieval, re-ranking, multi-document synthesis, and faithfulness evaluation
LangChain or LlamaIndex for orchestrating multi-step retrieval and generation workflows
SQL for document metadata management, evaluation dataset construction, and retrieval performance analysis
Experience with evaluation methodology for retrieval systems in specialised domains — precision at recall thresholds, attribution accuracy, and the standards appropriate to safety-critical deployment

Benefits

$106,000 – $132,000 base salary
Hybrid — three days in our Toulouse R&D facility, two days remote — relocation support available
Access to one of the most comprehensive aerospace technical documentation corpora in the private sector
$2,000 annual research and conference budget
Work at the intersection of NLP engineering and aerospace safety — a combination that is genuinely rare