CAST Software Inc.

CAST, a Software Company based in Meudon ,  is the market leader in Software Intelligence.

Working at CAST R&D means being an important part of a highly-talented, fast-paced, multicultural and Agile team . 

Overview

We’re building the foundation to ground AI with AAA Software Intelligence — Aggregated,

Accurated, and Augmented — sourced from real-world software and technology projects. This

role goes beyond manual curation: it's about using AI to empower AI. You will leverage LLMs,

embeddings, and NLP tools to clean, enrich, and validate data, enabling AI systems and

autonomous agents to rely on it for training and contextual understanding.

Responsibilities

• Aggregate and structure data from software ecosystems (codebases, APIs, tickets,

documentation, architecture specs).

• Apply LLMs, embeddings, and NLP tools to automate: data cleaning, entity extraction,

metadata tagging, and semantic annotation.

• Build and maintain semantic pipelines for LLM fine-tuning and RAG (Retrieval-Augmented

Generation).

• Organize datasets into formats suitable for Agent-to-Agent (A2A) interactions: APIs, vector

DBs, knowledge graphs, etc.

• Collaborate with AI teams to evolve schemas, prompts, labeling strategies, and evaluation

data.

• Ensure strong data lineage, reproducibility, and version control.

Requirements

• 3+ years in data engineering, ML data ops, or structured data curation.

• Proficient in Python, with strong data pipeline skills (Pandas, PyArrow, regex, Airflow).

• Experience with LLMs or NLP tools (e.g., Hugging Face, spaCy, LangChain).

• Ability to use AI to clean, enrich, classify, and organize technical content.

• Strong understanding of tokenization, chunking, and model input preparation.

• Experience working with software project data: Git repos, APIs, technical documentation, etc.

Bonus Skills

• Knowledge of vector DBs (FAISS, Qdrant, Weaviate) or knowledge graphs (Neo4j, RDF,

SPARQL).