Optimizing Vector Search Latency in Multi-Tenant RAG Pipelines
An analysis of indexing strategies using FAISS and HNSW to reduce semantic search latency by 40% while maintaining strict data isolation in multi-tenant environments.
Hi, I’m Arpita Jaiswal. An AI Engineer specializing in Agentic workflows, scalable LLM infrastructure, and high-performance web applications.

I am an AI Engineer and Full Stack Developer with a Master of Computer Applications. I specialize in bridging the gap between foundational LLM research and production-grade web architectures.
My core focus areas include:
An analysis of indexing strategies using FAISS and HNSW to reduce semantic search latency by 40% while maintaining strict data isolation in multi-tenant environments.
Exploring stateful, cyclical graph architectures to seamlessly handle LLM API failures, rate limits, and contextual hallucination recovery in production systems.
● Online & Listening
Architecting Agentic AI workflows and presenting on modern AI ecosystems at professional tech summits across Noida and the NCR region.
Developed and deployed highly scalable solutions, including DocuMind and SkillMatch-AI, bridging the gap between machine learning and functional UIs.
Graduated with honors. Specialized in Artificial Intelligence, Data Structures, and Advanced Algorithm Design.
Achieved top departmental rankings and earned several awards throughout the program.
Whether you need a scalable Agentic AI system, a high-performance web architecture, or just want to talk tech—my inbox is always open. Let's engineer something extraordinary.