Lê Chí ĐạiSelected work from research, coursework, and team projects. Each page describes the problem, my contribution, and the results.
A multi-agent LangGraph backend that parses CVs, scores candidate-job fit, and verifies claims against public sources, with RabbitMQ/Celery workers.
My contributionOwned architecture, parsing/chunking, worker pipeline, and R2/pgvector storage in a two-person team.
Led the architecture and AI backend of an Android reading assistant for ADHD users: OCR, Vietnamese narration with voice cloning, and retrieval-based Q&A. NAVER Hackathon national finalist.
My contributionLed core AI/backend implementation and coordinated end-to-end integration in a four-person team.
A multi-agent SRE copilot for a streaming market-data platform, evaluated by injecting faults and scoring answers against SQL answer keys.
My contributionBuilt the Analyst and Operations agents, A2A delegation behind the MCP gateway, the fault-injection eval and the model benchmark; teammates built the batch warehouse and Airflow scheduling.
Training and evaluation pipelines for image, text, and multimodal classification: CNNs, ViTs, BiLSTM/BERT, and CLIP, with augmentation experiments and LoRA few-shot adaptation.
Team scopeTeam project spanning vision, language, and multimodal training, with published experiments and error analysis.
Co-built an OpenGL renderer for annotated driving datasets and benchmarked YOLOv8, RT-DETR, and Faster R-CNN on accuracy and latency.
My contributionBuilt rendering hooks, annotation exports, and GLSL shaders for RGB, depth, and segmentation outputs.
A Python tool that checks HCMUT course materials and grades for changes and sends Discord notifications.
An educational OS simulator with a CFS-style scheduler, red-black-tree run queue, and paging-based memory management.
Exploring image reconstruction through interpolation, iterative back-projection, and patch-based methods.
A team-built benchmark suite across tabular data, text, images, and speech, with configurable pipelines and visual experiment reports.
Spoken-digit recognition with Hidden Markov Models, MFCC features, and a Gradio demo. Team coursework for CO3117 at HCMUT.
