AI Knowledge Engineer & Content Architect

Role Summary

At Sunrise, we operate in one of the most dynamic and fast-evolving industries in the world — telecommunications. As Switzerland’s leading challenger brand, we continuously innovate to deliver outstanding solutions to our customers. But what truly sets us apart is not only our technology — it’s our people.

At Sunrise Portugal, we are building strong, future-focused teams that combine expertise, creativity and ambition. We believe in empowering our employees, investing in development, and creating an environment where ideas are valued and impact is visible.

We are looking for a specialized AI Knowledge Engineer & Content Architect to design, create, and manage the internal knowledge ecosystem. This is a hybrid role sitting at the intersection of Knowledge Management and AI Technical expertise. You will not just be documenting information for our agents; you will be architecting content for machine consumption. You must understand the 'ins and outs' of how LLMs ingest data—from chunking strategies and metadata tagging to semantic retrieval—ensuring our models have the right information, at the right time, with high accuracy.

Key Responsibilities

1. LLM-Native Content Design & Structuring

• Structure for Machines: Transform complex internal documentation into LLM-ready formats (e.g., Markdown, JSON, structured text) to maximize machine readability and minimize parsing errors.
• Advanced Chunking Strategy: Define and implement recursive, semantic, and fixed-size chunking strategies to ensure content fits within context windows while preserving logical coherence.
• Metadata & Ontology Engineering: Design rich metadata taxonomies and filtering schemas that allow the AI to accurately narrow down search spaces during the retrieval phase.

2. Retrieval-Augmented Generation (RAG) Optimization

• Retrieval Logic & Reranking: Work with the technical teams to optimize how the model retrieves data. Fine-tune content phrasing and keywords to align with semantic search behaviors and vector similarity.
• Hallucination & Gap Analysis: Restructure content to provide definitive 'ground truth' to avoid hallucinations or failures to identify missing or conflicting knowledge.
• Token & Context Efficiency: Ensure content is formatted to maximize information density, reducing token costs while providing the LLM with sufficient context for high-quality generation.

3. Knowledge Pipeline & Governance

• Lifecycle & Truth Management: Establish rigorous workflows to manage and ensure that as internal policies change, the knowledge base is updated instantly. Resolve contradictions between legacy and new data.
• Data Integration: Manage the internal wikis and databases to the vector database in a seamless, automated, and secure way.
• Access Control Mapping: Ensure that the knowledge structure respects internal data permissions, so the LLM does not surface restricted info to unauthorized users.

4. Cross-Functional Translation

• SME to AI Translation: Interview Subject Matter Experts (SMEs) and convert their tacit knowledge into explicit, structured logic and 'golden sets' that the LLM can use for reasoning.
• Continuous Feedback Loops: Establish loops between users and developers to monitor response quality and iteratively improve the underlying knowledge assets.

Required Skills and Qualifications

1. Technical AI Knowledge

• Understanding of LLM Architecture: Working understanding of RAG (Retrieval-Augmented Generation), context windows, embeddings, and the difference between lexical and semantic search.
• Data Formatting: Proficiency in Markdown, JSON, and YAML, and an understanding of how document structure (headers, lists, tables) impacts model attention.
• Search Technology: Familiarity with vector databases (e.g., Pinecone, Milvus, Weaviate) and knowledge of reranking or hybrid search techniques.

2. Knowledge Management & Strategy

• Information Architecture: Proven experience creating enterprise-grade taxonomies and knowledge graphs.
• Precision Technical Writing: Ability to write with extreme clarity to eliminate linguistic ambiguity that could lead to model misinterpretation.
• Data Hygiene: Obsessive focus on version control and maintaining a 'clean' dataset free of redundant or obsolete information.

3. Tools & Analysis

• Basic Scripting (Preferred): Familiarity with Python (Pandas/LangChain) or SQL to automate data cleaning or query vector stores.
• Evaluation Frameworks: Experience using RAG evaluation tools (e.g., RAGAS) or analyzing search logs to identify performance bottlenecks.
• Language: English (mandatory) and intermediate level of DE, FR or IT a PLUS.
• Some experience with Telco desirable.

What’s next?

If you are interested in this position and meet the above requirements, please send your application (CV and motivation letter) to recruitmentportugal@sunrise.net

Key skills/competency

• AI Knowledge Engineering
• Content Architecture
• LLM-Native Design
• RAG Optimization
• Knowledge Management
• Information Architecture
• Vector Databases
• Technical Writing
• Data Governance
• Python Scripting