Agent systems, AI infrastructure, and distributed data systems
Ruowang (Jackie) Zhang
Tech Lead Manager and Staff Engineer building evaluation infrastructure, retrieval systems, data platforms, and long-horizon agent capabilities.
Profile
I work at the boundary of agentic AI, retrieval, model evaluation, and large-scale data infrastructure. My recent work spans Gemini agent evaluation systems, multimodal document indexing and vector search, and core storage and ingestion systems in Delta Lake and Databricks.
Recent Work
Selected from the most recent full-time roles.
Google DeepMind
May 2026 – Present
Mountain View, CA
Tech Lead Manager / Staff Engineer, Agent Systems
- Lead agent quality evaluation and infrastructure work for Gemini, including harnesses, data pipelines, and evaluation systems for agentic tasks.
- Collaborate with post-training and context engineering research teams on ultra long-horizon task performance, with emphasis on search and retrieval.
Contextual AI
Jul 2025 – May 2026
Mountain View, CA
Tech Lead Manager / Staff Engineer, AI Data Systems
- Led the data platform team across multimodal document parsing, embedding, indexing, vector search, and model inference infrastructure.
- Partnered with research on dynamic and agentic index building, model evaluation, reranking, and end-to-end RAG performance.
Databricks
Aug 2021 – Jul 2025
Mountain View, CA
Senior Software Engineer, Distributed Data Systems
- Built core Delta Lake abstractions and protocols for cost-free schema updates and large-scale customer data ingestion.
- Served as a top open-source Delta Lake contributor, contributed to Apache Spark, and led schema inference/evolution and file-source connector work across ingestion surfaces.
Research & Education
Purdue University – Doctor of Engineering in ECE, part-time/online. Research interests: agentic LLMs, RAG, RL, and context management.
Columbia University – M.S. in Computer Science, Machine Learning track. University of Michigan – B.S. in Computer Science and Data Science, Summa Cum Laude.
Recent research interests include context engineering, multi-agent systems, multimodal retrieval, and systematic model fine-tuning.
Technical Focus
AI Systems
Agent evaluation, RAG, retrieval, embeddings, reranking, multimodal document parsing, LLM tooling.
Data Infrastructure
Delta Lake, Apache Spark, ingestion systems, schema evolution, vector search, model serving infrastructure.
Languages
C++, Java, Scala, Python, Go, Swift, Objective-C, P4, SQL, TypeScript.
Interested in AI systems, retrieval, or data infrastructure?
Reach me at jackie@reactive.xyz.