📁 About our client:

Our client is the global leader in regulatory and sustainability intelligence, helping the world's largest companies navigate Environment, Health & Safety, Corporate Sustainability, and Product Compliance.

They are reinforcing their AI capabilities to keep their leadership as an AI-native compliance intelligence platform, with a foundational AI platform built around ontology, data, and platform surfaces, strong CAIO sponsorship, and direct executive air cover for the product layer of the AI transformation.

🚀 Responsibilities:

Data Platform & Retrieval

• Design, build, and own the Client & AI Data Vault end-to-end: document ingestion, clause-level parsing, chunking, versioning, and the query surface that every AI product consumes.

• Design the data model that maps client artifacts to the regulatory ontology so applicability and gap-analysis products can actually answer client-specific questions.

• Own the general database layer: SQL schema design, performance tuning, and the contract between operational stores and the AI retrieval surface.
  

Vector DB & RAG Operations

• Operate the vector database tier in production: indexing strategy, embedding model selection, similarity search tuning, recall vs. latency trade-offs, and the RAG infrastructure on top.

• Prototype chunking and embedding strategies rapidly against real client documents before committing to production patterns.

• Carry strong opinions (loosely held) about where embeddings earn their keep versus where classical retrieval still wins.
  

Document Intelligence & Parsing

• Build robust parsers for messy source formats: PDFs, DOCX, spreadsheets, scanned documents, regulatory filings, with clause extraction, table handling, and layout awareness.
  

Data Quality & Governance

• Define and enforce data quality: ingestion validation, embedding drift detection, versioning, and the audit trail every regulated AI product needs.
  

Team & Technical Leadership

• Lead a small squad of data and document intelligence engineers as a player-coach: roughly half of your time on hands-on build, half on direction-setting, review, and growing the people around you.

• Own hiring into the team and set the technical bar: code review culture, pipeline patterns, testing strategies, and operational discipline.

👤 Profile sought:

Experience:

• Clear history of shipping data and retrieval systems into production with measurable impact. Prototypes that never left the lab don't count.

• Hands-on experience building RAG systems or hybrid retrieval (vector + keyword + graph), including the failure modes and how to debug them.

• Track record designing chunking strategies and embedding pipelines for real-world documents, especially for legal, regulatory, or technical content.

• Production experience with relational databases: schema design, query optimization, and migration discipline.
  

Technical skills:

• Deep, hands-on experience operating a vector database in production (Qdrant, Pinecone, Weaviate, pgvector, or equivalent). Index tuning, recall vs. latency management, and clear opinions on which tool to reach for when.

• Strong Python skills, plus comfort with at least one compiled or systems-level language when performance matters.

• Hands-on experience parsing messy PDFs, DOCX, and structured extraction. Familiarity with layout-aware models, OCR, or clause-level extraction tooling is a strong plus.

• Production experience with cloud platforms (Azure preferred, AWS or GCP a plus). Comfortable with containerization, CI/CD, and the operational side of data infrastructure.

• Active user of AI coding assistants, integrated into daily workflow, with a clear point of view on what they are good and bad at, especially for parser development and schema exploration.
  

Bonus:

• Experience with multi-tenant data architectures and tenant isolation patterns.

• Familiarity with Elasticsearch, OpenSearch, or comparable full-text search engines.

• Background in NLP, information extraction, or document understanding.

• Experience with Kafka or equivalent message brokers for ingestion pipelines.

• Experience in regulated industries (EHS, legal, financial, healthcare) where auditability and versioning are non-negotiable.

• Contributions to open-source retrieval, embedding, or document parsing tooling.
  

Languages:

• Fluent English required.
  

Soft skills:

• Player-coach leadership style: technically credible, willing to ship code, equally focused on growing the team.

• Opinionated about tools, pragmatic about deadlines.

• Strong communicator across technical and business audiences.

• Bias for delivery and operational discipline over research polish.

🌍 Benefits & Culture:

• Tech stack: Vector DBs (Qdrant / Pinecone / Weaviate / pgvector), Python, Azure, RAG, embeddings, document parsing, SQL, Kafka, Elasticsearch.

• Direct access to leadership and the Chief AI Officer: short feedback loops, real influence on architecture and direction.

• A seat on a small, high-impact AI team where the data layer is the single largest product-portfolio unlock.

• Culture that treats AI tools as force multipliers, not novelties.

• Competitive compensation, benefits, and flexibility.

• Hybrid in Lisbon's Office role (3 days a week at the office)
  
  

💼 Department: AI & Engineering

📍 Location: Lisbon

📆 Start date: ASAP