AI Engineer building practical systems across machine learning, computer vision, AI automation, and agentic AI. I focus on taking AI projects from data pipelines through model training to deployed applications and APIs.
My work spans:
- Computer Vision: Image classification, object detection, transfer learning, RGB/thermal analysis
- AI Automation: Workflow orchestration, tool-calling agents, automated pipelines
- ML Systems: End-to-end systems from data processing to model inference to APIs
- AI Data Operations: Data annotation, dataset preparation, quality control, model-assisted labeling
Currently building intelligent systems that combine LLM reasoning with deterministic tool execution for practical AI applications.
Agent architectures with tool calling, structured outputs, and deterministic execution pipelines. |
End-to-end CV pipelines from image classification to object detection to production APIs. |
RAG pipelines with source citations, hallucination detection, and evaluation metrics. |
Data annotation workflows with model-assisted labeling and quality validation. |
|
Retrieval-Augmented Generation system with multiple chunking strategies, hybrid search, reranking, hallucination detection, and evaluation metrics. 27 tests. Tech: Python, NumPy, FastAPI, OpenAI, pytest |
AI agent with tool calling, LLM-based planning, result validation, and code analysis. Supports streaming responses and structured outputs. 42 tests. Tech: Python, FastAPI, OpenAI, Pydantic, pandas |
|
Computer vision system for object detection, defect analysis, batch processing, and quality scoring with statistical reporting. 37 tests. Tech: Python, OpenCV, YOLO, FastAPI, NumPy |
Multi-agent AI system combining RAG chatbot with CNN vision for plant disease diagnosis. Features LangChain RAG pipeline and Flask inference API. Tech: Python, TensorFlow, LangChain, OpenCV, Flask, Streamlit |
|
TypeScript RAG client with strict typing, chunking strategies, vector operations, and evaluation metrics. Full test coverage. Tech: TypeScript, Vitest, OpenAI, zod |
Comprehensive code review guidelines with quality checklists, pattern catalogs, and anti-pattern detection. Practical examples for AI-assisted review. Tech: Markdown, Python, TypeScript |
All projects include unit tests, integration tests, CI/CD pipelines, and Docker containerization.
Building agentic AI systems with tool calling and workflow orchestration
Learning advanced RAG patterns and LLM evaluation
Developing computer vision + LLM multimodal systems
Exploring AI data operations and annotation pipelines
Building practical AI systems — from data to deployment