AI Evaluation & Retrieval Scientist (LLM Extraction)

Volta Software; Paris; Ibrido; EUR 70.000 - 110.000; Tempo pieno

Descrizione del lavoro

Volta Software seeks an Applied Data Scientist who measures and improves accuracy across orders. You will define per-field metrics, build automated evaluation harnesses and enable scalable, configurable prompts for extraction and matching using LLMs.

You’ll work with Python, SQL and Elasticsearch to handle messy documents, PDFs, and multi-language data, collaborating with enterprise customers to drive measurable improvements.

Competenze

• 3+ years shipping ML or NLP to production.
• Strong Python, Pandas or Polars and testing habits.
• Hands-on with LLM APIs (Gemini, OpenAI, Anthropic or similar).
• Experience with Elasticsearch or similar search and retrieval.
• Experience with evaluation metrics (precision/recall, per-field accuracy).
• Fluent English; French or Italian for documents.

Mansioni

• Define how accuracy is measured per field and track per customer/release.
• Build evaluation harnesses and CI-integrated datasets.
• Improve matching with catalogues and synced master data.
• Turn customer fixes into reusable prompts/configuration templates.
• Perform rigorous error analysis to identify root causes and fixes.
• Turn data issues into structured evaluation outputs.

Conoscenze

Python

Pandas

Polars

Testing habits

LLM APIs

Elasticsearch

SQL

Error analysis

English fluency

French or Italian

Formazione

MSc in CS/Stats/Applied Math

Strumenti

OpenSearch

Langfuse

OCR tooling

GCP

Descrizione del lavoro

Volta Software seeks an Applied Data Scientist who measures and improves accuracy across orders. You will define per-field metrics, build automated evaluation harnesses and enable scalable, configurable prompts for extraction and matching using LLMs.

You’ll work with Python, SQL and Elasticsearch to handle messy documents, PDFs, and multi-language data, collaborating with enterprise customers to drive measurable improvements.