Scaling AI Agents for Production Codebases: Patterns for Accuracy and Efficiency
watch on portal.gitnation.org ↗https://portal.gitnation.org/contents/scaling-ai-agents-for-production-codebases-patterns-for-accuracy-and-efficiencysummary
Shows three concrete architectural patterns—Language Server Protocol (LSP)–based semantic code intelligence, bundling focused agent skills with their required context and tools, and delegating work to parallel subagents—to enable AI agents to handle large production codebases. Demonstrations use a real open-source project (example: ShadCN) to illustrate multi-file refactorings, cross-module changes, and dependency tracking, and explains how these techniques avoid context-window overload and outperform plain text search.
AI-written summary · about
related talks
40mBuilding Effective AI Agents Simple, Composable Patterns
59mFrom Pair to Peer Programmer
29mPairing with the machine: How AI Agents are rewriting the rules of software development
01h 03m10 tips to level up your ai-assisted coding
34mAI Agents: Transforming Autonomous Decision-Making
43mArchitecting the Future: Modern Solution Design with Agentic AI
30mBuilding Tools with Tools: Modern AI-Driven Development
44mBuilding Agentic AI Systems: From Transactions To Conversations