Julio Sanchez

Data Research Analyst @Spotwork

Toronto, ON, CA

Developed and deployed Conversational AI Bot API leveraging VertexAI LLM on Google Cloud Platform, prototyping intelligent components for advanced conversational interfaces.Implemented and maintained Conversational AI applications in a production environment, enabling seamless voice and text interactions with users.Led the integration of CircleCI for streamlined DevOps and automated pipelines in a Python environment, enhancing deployment efficiency.Utilized the LangChain Python framework with techniques such as RAG, Chain of Thought, and Custom Agents, achieving a 92% accuracy rate in decision-making processes.Experimented with prompt engineering techniques including Chain of Thought, significantly refining LLM decision-making and improving conversational abilities.Reduced customer interactions by 80%, achieving an average response time of 3 seconds for user queries, enhancing user satisfaction.Employed Firebase for secure NoSQL storage and retrieval of conversation data, ensuring data integrity and security.Enhanced system performance with ElasticSearch, utilizing it for vector stores in the RAG system, thereby optimizing query handling and response generation.Conducted research on the latest advancements in NLP/NLU and language models (LLMs) to continually improve chatbot capabilities.Collaborated frontend and backend teams, contributing to the multidisciplinary design and implementation of conversational agent workflows.Engaged in rapid prototyping and experimentation, advising on best practices for designing and integrating conversational agents, continually refining the AI system\'s capabilities.

Pontificia Universidad Javeriana

Grado en Ingeniería, Ingeniería eléctrica, electrónica y de comunicaciones

Lambton College

Artificial Intelligence & Machine Learning Post graduate Program, Ingeniería informática