AI Data & Architecture Specialist

AI Data & Architecture Specialist Location: Auckland, New Zealand (Hybrid/Remote options available) Department: AI Operations Reports to: Team Lead – AI Operations The Mission: Architect the Brain of our Autonomous Workforce At Straker, we aren't just using AI; we are building a sophisticated, autonomous workforce. As our AI Data & Architecture Specialist , you are the custodian of the "Knowledge Layer." You understand that an AI agent is only as smart as the data it can access. Your mission is to ensure our agents have access to high-quality, structured, and contextually relevant data. You won’t just be managing a database; you will be designing the data pipelines, vector stores, and retrieval architectures (RAG) that allow our AI to make informed, "needle-in-the-haystack" decisions. What You’ll Be Doing 1. Knowledge Architecture & RAG Strategy Design and optimize Retrieval-Augmented Generation (RAG) pipelines to provide agents with pinpoint-accurate context. Own the selection and scaling of Vector Databases (Pinecone, Weaviate, Milvus). Develop sophisticated metadata schemas to improve search relevance and agent "retrieval precision." 2. Data Engineering for AI Build ETL pipelines that turn messy, unstructured data (PDFs, emails, docs) into "LLM-ready" formats through smart chunking and embedding. Lead the creation of synthetic datasets for agent training and benchmarking. Establish rigorous cleaning protocols to eliminate stale or biased information. 3. Model & Embedding Selection Evaluate and select embedding models for specific domains (Legal, Finance, Technical). Collaborate on model fine-tuning when off-the-shelf LLMs aren't enough. Continuously tune hybrid search (keyword + semantic) to optimize performance. What Success Looks Like We measure impact through precision and speed. In this role, you’ll be aiming for: Accuracy: A 90% "Hit Rate" for agent context retrieval. Freshness: Ensuring the Knowledge Base updates within 60 minutes of a source change. Performance: Optimizing vector search queries to return results in under 250ms. Scalability: Building an architecture that handles data growth without linear cost increases. Your Technical Profile The Must-Haves: Data Engineering: Expert-level Python and SQL . Experience with orchestration tools like Airflow or Dagster. Vector Mastery: Deep experience with Vector DBs and similarity search algorithms (HNSW, IVF). NLP Fundamentals: A strong grip on tokenization, embeddings, and semantic chunking. Cloud Infrastructure: Experience managing large-scale data on AWS, GCP, or Azure. The Nice-to-Haves: Experience with Graph Databases (e.g., Neo4j) for relationship mapping. A background in Information Retrieval (IR) or Library Science. Familiarity with AI data privacy frameworks (GDPR/NZ Privacy Act). Are You the Ideal Candidate? You are obsessed with "Data Truth." You enjoy the challenge of turning messy, unstructured human knowledge into a clean, searchable, and machine-readable architectural masterpiece. You are organized, analytical, and ready to build the future of AI-driven operations. Ready to architect the future? Apply now.