# KDnuggets Weekly Roundup: Week of July 13, 2026 > Source: > Published: 2026-07-18 13:00:05+00:00 # KDnuggets Weekly Roundup: Week of July 13, 2026 Stop Using If-Else Chains: Use the Registry Pattern in Python Instead • 5 Real-World SQL Projects to Build Your Data Portfolio • 10 YouTube Channels Keeping You Ahead in AI • Structured Language Model Generation with Outlines **🐍 Stop Using If-Else Chains: Use the Registry Pattern in Python Instead** Kanwal Mehreen · Python · July 15, 2026 Long conditional chains hinder extensibility in Python by violating the Open/Closed Principle, making code brittle when new options are introduced. The Registry Pattern solves this by replacing hardcoded dispatch logic with a central lookup table where components register themselves dynamically. Implementing this pattern allows system behavior to be driven by configuration, resulting in more maintainable and easily extensible pipelines. ➡️ **12 Ways to Reduce LLM Latency and Inference Costs in Production** Kanwal Mehreen · Language Models · July 14, 2026 Reducing LLM latency and inference costs in production requires optimizing workflow design by minimizing token usage, employing model routing for specific tasks, implementing multi-layered caching strategies, and managing context budgets rather than relying solely on larger contexts or aggressive batching. ➡️ **5 Real-World SQL Projects to Build Your Data Portfolio** Abid Ali Awan · SQL · July 13, 2026 Building a strong data portfolio requires executing real-world SQL projects across domains like customer churn, data warehousing, sales analysis, banking segmentation, and healthcare to demonstrate the ability to clean data, model systems, and derive actionable business insights. ➡️ **Git Worktrees for AI Development** Shittu Olumide · Programming · July 17, 2026 Git worktrees provide an essential infrastructure layer that enables multiple AI agents to operate simultaneously on a single repository by creating isolated workspaces, eliminating the risk of file collisions and context loss during parallel development. ➡️ **Structured Language Model Generation with Outlines** Iván Palomares Carrascosa · Language Models · July 13, 2026 The Outlines library introduces deterministic certainty into LLM output generation by masking syntactically illegal tokens, enabling practitioners to reliably obtain strictly structured outputs like JSON by enforcing specific constraints during inference. ➡️ **7 Python Frameworks for Orchestrating Local AI Agents** Shittu Olumide · Artificial Intelligence · July 15, 2026 Seven Python frameworks provide the necessary orchestration layers for building, coordinating, and running secure, cost-effective AI agents directly on local infrastructure. ➡️ **10 YouTube Channels Keeping You Ahead in AI** Vinod Chugani · Artificial Intelligence · July 16, 2026 A curated selection of ten YouTube channels provides comprehensive, high-quality educational content spanning machine learning theory, deep learning implementation, paper analysis, LLM application development, and industry trend tracking for accelerating professional AI knowledge. ➡️ **Getting Started with Conductor for Gemini CLI** Shittu Olumide · Programming · July 14, 2026 Conductor introduces Context-Driven Development to resolve context issues in AI coding by persisting project specifications and architectural context in repository files, enabling agents to generate accurate code based on established project constraints across sessions. ➡️ **5 FREE Resources on Agentic AI** Nahla Davies · Artificial Intelligence · July 17, 2026 A curated set of free resources provides a structured path for practitioners to move beyond building agent demos by integrating hands-on framework experience, theoretical foundations in multi-agent systems, orchestration patterns, and essential evaluation techniques. ➡️ **Working with Pi Coding Agents** Shittu Olumide · Programming · July 16, 2026 Pi Coding Agents advocates for a minimalist architectural approach by explicitly documenting the features it omits, arguing that reducing built-in complexity and injected context leads to more efficient and cost-effective agentic workflows.