Vetted Vector Search Professionals

“Entrepreneurial builder and former founding CTO of a Y Combinator startup that was acquired. Now in the planning stage for a new venture (considering edtech, media creation, and food/"intelligent servicing"), emphasizing early design partners, rapid testing, and iteration to validate ideas.”

“Built a production LLM-powered root cause analysis agent for supply chain alerts that helped operations managers avoid manual dashboard investigation. Demonstrates unusually strong depth in agent reliability, orchestration, and observability, with concrete production practices like hallucination blocking, shadow testing on 500 cases, and data-driven improvements that raised user agreement to 94% while cutting GPT-4 usage by 60%.”

“AI/LLM engineer with production experience at NVIDIA and Microsoft, including building a RAG-based enterprise knowledge assistant that improved accuracy by 42% and scaled to thousands of queries. Deep in inference optimization (TensorRT-LLM, Triton, quantization, speculative decoding) and MLOps/observability (Prometheus/Grafana, MLflow, LangSmith), plus orchestration with Kubeflow/Airflow across multi-cloud.”

Yishi Wang

Screened

Junior Machine Learning & Data Science professional specializing in LLMs and analytics

Chicago, IL3y exp

“Amazon internship experience building production GenAI analytics for the returns organization: a multi-agent LLM+RAG system that let analysts query multiple heterogeneous data sources in natural language without hand-written SQL. Also built and operationalized four Apache Airflow DAGs for large-scale ETL, emphasizing observability and freshness-aware metadata to keep outputs accurate and up to date.”

Vinay Ramrupe

Screened

Mid AI/ML Engineer specializing in LLM and enterprise generative AI

San Francisco, CA5y exp

“ML/AI engineer focused on taking LLM systems from experimentation to reliable production, including enterprise copilot and RAG-based knowledge retrieval use cases. Stands out for combining data pipelines, model training, inference optimization, automated evaluation, and safety guardrails, with cited impact including 20% throughput gains and 30% less manual evaluation effort.”