Thoughts on AI systems, agent architecture, and production engineering

A controlled experiment testing whether adding purpose to LLM system prompts changes decision-making behavior. Tested across 3 model families, validated with psychometric methodology, and confirmed in production.

Learn how I built Vectus AI, an advanced medical scheduling assistant powered by OpenAI GPT-4. This technical deep dive explores the architecture, implementation, and challenges of creating an AI-driven conversational agent for streamlining medical appointments.

Explore the technical journey of creating Q Bot, an AI-powered Twitter bot that generates Star Trek Q character-themed content using Node.js, OpenAI GPT, and the Twitter API.

A technical exploration of building an AI-powered bot to detect and explain logical fallacies in social media posts using Python, multiple LLM models, and web automation.