AI Healthcare
AI Developer, Apr 2023 - Present
Project description
- Voice Bot Assistance: Engineered a voice-activated bot to assist doctors and nurses with patient management and task organization, improving workflow efficiency and reducing administrative burden.
- Natural Language Processing (NLP) Application: Implemented an NLP application capable of autonomously generating clinical documentation, such as Diagnostic Assessment and Plan (DAP) notes and Progress notes, thereby minimizing manual input and potential errors.
- AI-Powered System: Created an intelligent chat platform that aids case managers in appointment scheduling. The system includes a predictive feature that assesses the likelihood of patient no-shows for upcoming appointments, enabling proactive measures to optimize resource allocation and enhance appointment adherence.
Responsibilities
- Develop and train pre-trained NLP models: This involves designing and training deep learning models on large text datasets for specific NLP tasks.
- Utilize Azure AI Services for NLP functionalities: Leverage Azure AI features like Speech Transcription and Meeting Summarization to enhance the solution.
- Construct data pipelines with Azure Data Factory: Design and implement automated data pipelines using Azure Data Factory for efficient data processing and report generation.
- Implement core algorithms for automated clinical document generation: Develop the algorithms responsible for generating clinical documents based on extracted data.
- Maintain and support the solution: Implement a maintenance plan for continuous improvement, bug fixing, and assisting team members.
Technologies
- Azure AI Services, Azure Data Factory, Azure Storage, OpenAI, Langchain, Prompt Engineering, Azure DevOps.
NLP Toolkit
AI Developer, Oct 2022 - Apr 2023
Project description
- Working on a team of 3 members including 1 Team Leader and 2 Developers.
- NLP Toolkit is an app built specifically for Natural Language Processing. We provide an extremely easy API to quickly apply pretrained NLP models to text data, such as Text Summarization, Sentence Similarity, Grammar Correction, etc.
Responsibilities
- Build the pretrained models for NLP tasks.
- Setup and deploy pretrained models on the server
- Improve the pretrained model that fits each NLP task.
- Fixing bugs, supporting team members and maintaining.
Technologies
- PyTorch, Docker, HuggingFace.
FaceAttendance – Internal use
AI Developer, Oct 2022 – Present
Project description
- Working on a team of 3 members including 1 Team Leader and 2 AI Developers and 2 Web Developers.
- Face Attendance is a system that supports tracking attendance of the internal employee.
Responsibilities
- Research methods improve face recognition models.
- Setup and deploy code on the server.
- Implement testing methods for face recognition models.
- Review code and develop the features.
Technologies
- PyTorch, Docker, Onnx, PostgreSQL
ChatBot - Japanese
AI Developer, Apr 2022 – Oct 2022
Project description
- Working on a team of 5 members including 1 Project Manager, 1 Team Leader, 2 AI Developers and 1 Manual QA.
- Building a Chatbot for Japanese clients using Amazon Lex Services, supports multiple platforms (MsTeams, RocketChat, Slack).
Responsibilities
- Make the technical proposal.
- Proposing methods for improving the performance of Chatbot.
- Implement support modules for the chatbot, like Japanese Tokenizer (SudachiPy, Juman++).
- Implement retrain algorithm Amazon Lex model.
- Write documents and daily reports for customers.
Technologies
- PyTorch, AWS Lex, AWS DynamoDB, AWS Lambda Function, S3 Bucket.
Smart E-learning – Japanese
Developer, Nov 2021 – Oct 2022
Project description
- Smart E-learning is a smart learning management system that allows targets like educational centers which can be customized according to personal positions such as Learners, Teachers or trainers, and Administrators.
- Working on a team of 10 members including 1 Project Manager, 1 Team Leader, 3 Developers and 5 QA.
Responsibilities
- Implement the algorithm to parse the exam file to components such as header, questions, answer and result, etc.
- Support my colleagues implement the algorithm that auto checks and gets the score of the assignment of the students using YOLOv5.
- Implement the algorithm that auto generates new exams from a database of parsed questions.
Technologies
- PyTorch, YOLOv5, FastAPI, OracleSql.
ChatBot - Japanese
Developer, Dec 2021 – Apr 2022
Project description
- Working on a team of 3 members including 1 Team Leader, 2 Developers.
- A Chatbot for internal employees using RASA Chatbot Framework, supporting multiple platforms (Odoo, RocketChat).
Responsibilities
- Make the technical proposal.
- Create a script for generating data for train Chabot (using Chattete).
- Proposing methods for testing and improving the performance of Chatbot.
- Implement Vietnamese language model to Chatbot architecture.
- Write a dockerfile and deploy the chatbot on an EC2 instance.
Technologies
- PyTorch, RASA, Microsoft Bot Framework, AWS EC2
