NLP/LLM/Conversational AI Expert

James Smith

AI/Chatbot/NLP Expert

***************@*****.*** Toronto, Canada

Profile

An experienced NLP expert with a 9-year track record of creating innovative software solutions that convert large text datasets into structured knowledge graphs. Skilled in developing smart chatbot systems and utilizing sophisticated vectorization to enhance machine-human language interaction. Passionate about leveraging machine learning and NLP for actionable insights from unstructured data, with a dedication to advancing AI and computational linguistics. Professional Experience

Senior AI/NLP Engineer, Eddy Travels

•Spearheaded the design and deployment of an custom chatbot using LLM, integrating with Azure SQL achieving a 40% reduction in customer support workload by automating responses for fundraising participants.

•Utilized Word2Vec for converting extensive customer logs into actionable data, enhancing sentiment analysis by 30% and guiding product development strategy.

2018 Apr – present

Toronto, Canada

•Enhanced the chatbot LLM's language modules using RAG tools, yielding a 50% improvement in query resolution and a 20% decrease in live support dependency.

•Developed an NLP-based rating system using embeddings and vectorization, processing user reviews and enhancing customer satisfaction metric precision by 35%.

NLP Engineer, Rarelogic

•Led development of advanced sentiment analysis algorithms, optimizing feedback processing and boosting customer satisfaction scores by 25%.

•Designed and trained machine learning models for entity classification, improving data categorization accuracy by 30% and operational efficiency in marketing and sales.

2016 Feb – 2018 Mar

Ottawa, Canada

•Developed an NLP engine for text summarization, automatically generating product descriptions from catalog data, reducing content production time by 40%.

•Developed a personalized recommendation system with collaborative filtering, achieving a 15% uplift in cross-sales and 20% more repeat purchases.

Junior AI Engineer, A.S. Watson Group

•Developed and implemented TensorFlow image processing pipeline, increasing pattern recognition accuracy by 35% on financial datasets.

•Designed RNN-based translation model, reducing errors by 25% and speeding processing by 40%, enhancing multilingual document analysis. 2014 Jan – 2016 Jan

Hong Kong

•Engineered a CNN image classification system, achieving 92% precision in identifying financial chart patterns.

•Worked collaboratively with data scientists and engineers, contributing to a 15% increase in the efficiency of the company’s AI-driven trading strategies. ***************@*****.*** 1 / 2

Education

Bachelor's Degree of Computer Science, Hong Kong Institute of Technology GPA: 3.9/4.0

2010 Sep – 2013 Dec

Hong Kong

Skills

Machine Learning

Data Analysis, Time Series Prediction, Image

Classification, Recommendation System

Frameworks

Python, PyTorch, transformers, Tensorflow,

NLTK, SpaCy, R, SQL, AWS, GCP, Azure, Azure

SQL, FastAPI, Flask

Natural Language Processing

LLM development, OpenAI GPT, LLama, Mixtral,

Chatbot Development, Algorithm Development,

Model Training, Text Processing, Speech

Recognition, Sentiment Analysis, Name Entity

Recognition, RAG tools, Sentence Vectorization,

Knowledge Graph Construction, Information

Retrieval

Projects

SalesAssistantBot, Sales Bot in Car Seller Site

Integrating Sales Chatbot in Car seller site using RAG tools for recommendation customization.

https://www.tmcmotors.co.uk/search/

2023 Jul – 2024 May

HealthCareChatbot

Developed HealCareChatbot using GPT-4o adding functionalities to gathering healthcare information from MSD dataset for improving accuracy while inserting hand-written prescription images for chatbot customization. https:/talkhealth.ai/main

2023 Jun – 2024 May

GeoSearchEngine, Search Engine for Goolemaps using Natural Language Developed GoogleMaps search engine using Natural Language integrating with RAG tools for adding External Database to the GPT bot. https://hunter-project.vercel.app/ (username: test, password: productionglue)

2023 Oct – 2024 Mar

***************@*****.*** 2 / 2

Contact this candidate