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
Implement a text-to-SQL system using Anthropic's Claude API with tool use for SQL execution, multi-turn conversations, and structured output parsing for reliable database question answering.
Build a text-to-SQL system with automatic error detection, self-correction loops, and intelligent retry strategies that fix query mistakes without user intervention.
Understand how to evaluate text-to-SQL systems using the Spider and BIRD benchmarks, implement execution accuracy metrics, and build custom evaluation datasets for your specific database schema.
Learn what text-to-SQL is, how the architecture works from schema understanding to query generation, and why it is one of the most practical applications of large language models in enterprise software.
Compare open-source text-to-SQL models including SQLCoder, NSQL, and DeFog SQLCoder. Learn how to deploy them locally, fine-tune on your schema, and evaluate accuracy against commercial alternatives.
Master the decision framework for when AI agents should reach for external tools versus relying on pure reasoning, with practical heuristics for tool selection, hybrid approaches, and cost-benefit analysis.
Comprehensive guide to every action type UFO can perform — from basic clicks and keyboard input to scroll operations, UIA element interactions, and application-specific control manipulation.
Deep dive into Microsoft UFO's dual-agent system where HostAgent orchestrates application selection and AppAgent executes in-app UI actions, with detailed coordination flow and plan execution examples.
Understand Microsoft UFO's known limitations with complex UI controls, high-DPI displays, and time-sensitive interactions, along with practical workarounds and hybrid strategies for production reliability.
