Zipporin is seeking an experienced Principal Architect to design, scale, and operationalize our next-generation Feature Store and Vector Database platforms. Based out of our Seattle office on a hybrid model, you will lead the architectural strategy bridging high-throughput distributed data systems with modern enterprise AI/ML and retrieval-augmented generation (RAG) workloads.
Key Responsibilities:
- Architect, design, and govern low-latency, highly scalable feature store and vector database infrastructure supporting production real-time inference and batch training.
- Evaluate, benchmark, and deploy enterprise-grade vector indexing engines and similarity search solutions (e.g., Milvus, Pinecone, Qdrant, pgvector).
- Collaborate with ML engineers and data platform teams to optimize feature definition, entity consistency, point-in-time correctness, and offline-to-online feature synchronization.
- Establish data governance, access controls, lineage, and lifecycle management policies across multi-tenant vector embeddings and feature sets.
- Drive capacity planning, reliability engineering, and performance optimization for mission-critical ML data platforms operating at petabyte scale.
Qualifications & Requirements:
- Proven track record as a distributed systems architect or principal engineer building ML infrastructure, real-time data pipelines, or high-dimensional search platforms.
- Deep technical expertise with feature store platforms (such as Feast, Tecton, or Hopsworks) and vector databases.
- Proficiency in streaming and batch technologies, including Kafka, Flink, Spark, and modern cloud data warehouses.
- Strong background in systems programming and automation using Python, Go, Java, or C++, alongside container orchestration via Kubernetes.
- Bachelor's or advanced degree in Computer Science, Computer Engineering, or equivalent professional experience.