Agentic AI, LLM engineering, and the models behind modern automation — multi-agent systems, LLM evaluation and comparisons, RAG, fine-tuning, AI infrastructure, security, and production AI engineering.
From the blog
Build conversation summarization systems that generate concise, actionable summaries of long AI agent interactions with key point extraction, decision tracking, and follow-up items.
Build a conversational booking agent that integrates with calendar APIs, handles timezone conversions, checks real-time availability, and manages confirmation and reminder flows.
Learn how coreference resolution enables AI agents to track pronouns and references across conversation turns, with practical implementations using spaCy, neural models, and LLM-based approaches.
Learn how Corrective RAG (CRAG) adds relevance scoring, re-retrieval, and web search fallback to catch and fix bad retrievals before they reach the user. Full Python implementation included.
Implement proven strategies to dramatically reduce AI agent operating costs through intelligent model routing, response caching, request batching, prompt optimization, and usage caps without sacrificing user experience.
Learn to accurately track and optimize the total cost of AI agent conversations including token usage, tool call expenses, infrastructure overhead, and strategies for reducing cost per interaction.
Build a complete cost tracking system for AI agents that attributes token usage to individual users and features, sets budget alerts, and provides dashboards for controlling LLM spend in production.
Master the art of designing effective CrewAI agents by crafting specific roles, meaningful backstories, aligned goals, and configuring verbose mode for transparent agent reasoning.
Implement step callbacks, task callbacks, and custom event handlers in CrewAI to monitor agent reasoning in real time, log progress, and build observable multi-agent systems.
