My Tech Stack Evolution

Over the past three decades I moved together with the industry, adopting each generation of tools as it matured. Here are some technologies I have used extensively across industry roles, applied research, and teaching.

Now GenAI & Agents

The GenAI Epoch

Building with large language models, retrieval-augmented generation, tool-using agents, and orchestrated multi-agent systems.

Foundation/Diffusion/Multimodal Models Transformers OpenAI Hugging Face LangChain/LangGraph RAG/Vector DB Multi-Agent (CrewAI) MCP Vibe-Coding
2020s ML at Scale

Deep Learning and Data Science at Scale

Distributed training on Spark clusters, model lifecycle management with MLflow, and production deployment across AWS and Azure.

Databricks/PySpark SparkNLP/SpaCy Horovod VAE/GAN Architectures Graph ML (StellarGraph/Office Graph) MLflow Feature Store ONNX AWS/Azure
2010s ML & Deep Learning

Transition from Classical ML to Deep Learning

Moving from statistical models and gradient boosting to neural networks, applying both to finance, public safety, and automotive domains.

Python R KDB+ NumPy/Pandas/SciPy scikit-learn XGBoost LightGBM TensorFlow PyTorch Keras CNN/RNN Architectures
2000s Multimedia & Vision

Computer Vision and Multimedia for Video Streaming and Client-Server Architectures

Integrating real-time video capture, codec pipelines, and vision algorithms into client-server systems for education, content delivery, and surveillance.

OpenCV IPP DirectShow Video Streaming H.264 MPEG-4 RTP/RTSP/WMS VoIP ASP HTML/JavaScript
1990s Systems & Algorithms

Native Algorithm Development

Algorithm design in C++ for signal processing, handwriting recognition, and network protocols on desktop and embedded platforms.

C++ (Borland/Visual) MATLAB SQL/ODBC Win32/MFC COM/DCOM ISAPI TCP/IP WinSockets/NDIS PalmOS

Quick Links