Valerio Valente

Developed a RAG-based retrieval agent leveraging AWS and vector search to automate document analysis workflows. Conducted vector database operations for knowledge indexing and semantic retrieval. Delivered BI dashboards in Apache Superset, improving decision-making efficiency for manufacturing clients.

Designed and deployed LLM-based applications, including a multi-agent system with a Model Context Protocol (MCP) server. Generated and curated evaluation datasets for internal LLM benchmarking and improvement. Contributed to the mcpadapt open-source project, improving developer tooling for LLM adaptation workflows. Recognized as a top 5 performer in a FDA’s challenge, representing InSilicoTrials and demonstrating excellence in deploying AI solutions under regulatory standards. Built and deployed LLMOps pipelines enabling GenAI for tasks like optimization-by-prompting (OPRO), document parsing, and code generation. Collaborated in an agile environment, mentoring peers and conducting technical interviews.