AI Knowledge Engineer & Content Architect - Full Remote Portugal

ABOUT THE OPPORTUNITY

A rare and genuinely cutting-edge role at Switzerland's leading telecommunications challenger brand, now building out their Portugal hub. This isn't a standard knowledge management or technical writing position — it sits at the frontier of AI engineering, where your work directly determines how well LLMs reason, retrieve, and respond. If you understand how machines consume information and want to architect that process from the ground up, this is your opportunity.

Engagement type: Direct Hiring | Level: Mid-level | Base salary: Up to €45,000 (total compensation up to ~€50,000 with bonus and perks) | Language: English fluent (required) | Location: Portugal-based (Lisbon area preferred)

PROJECT & CONTEXT

You'll be the architect behind an internal AI knowledge ecosystem — designing how information is structured, chunked, tagged, and retrieved by LLMs in a production RAG environment. Working at the intersection of knowledge management and AI engineering, you'll collaborate with SMEs, technical teams, and business stakeholders across Portugal and Switzerland to ensure the organisation's AI agents have the right information, at the right time, with high accuracy. Full remote, with occasional in-person visits to the Lisbon office for team events or collaborative sessions.

WHAT WE'RE LOOKING FOR

• Working understanding of LLM architecture concepts — particularly RAG (Retrieval-Augmented Generation), context windows, embeddings, and the distinction between lexical and semantic search
• Hands-on experience designing and implementing chunking strategies (recursive, semantic, fixed-size) to preserve logical coherence within context window constraints
• Proven ability to design metadata taxonomies, ontologies, and filtering schemas that improve AI retrieval accuracy and narrow search spaces effectively
• Proficiency in Markdown, JSON, and YAML, with a clear understanding of how document structure (headers, lists, tables) influences model attention and parsing quality
• Familiarity with vector databases such as Pinecone, Milvus, or Weaviate, and knowledge of reranking or hybrid search techniques
• Experience in information architecture — creating enterprise-grade taxonomies, knowledge graphs, or structured content systems at scale
• Demonstrated ability to conduct hallucination and gap analysis — identifying missing, conflicting, or ambiguous knowledge and restructuring content to provide reliable ground truth
• Strong technical writing skills with extreme clarity and precision, eliminating linguistic ambiguity that could lead to model misinterpretation
• Experience establishing knowledge lifecycle workflows — managing version control, resolving contradictions between legacy and updated data, and maintaining a clean, non-redundant dataset
• Ability to interview SMEs and translate tacit knowledge into structured, explicit logic usable by LLMs ("golden sets")
• Fluent in English — mandatory for all stakeholder communication and content creation

NICE TO HAVE

• Basic scripting skills in Python (Pandas, LangChain) or SQL for automating data cleaning or querying vector stores
• Experience with RAG evaluation frameworks such as RAGAS, or with analysing search logs to identify retrieval performance bottlenecks
• Proficiency in German, French, or Italian — a meaningful advantage given the Swiss operational context
• Exposure to access control mapping and data permission structures within AI knowledge systems
• Background in Telco or other complex, regulated industry environments
• Experience building continuous feedback loops between end users and development teams to iteratively improve knowledge assets