Sub-Agents in Claude Code 2.1: When and How to Spawn Them
A practical engineering deep dive into Claude Code 2.1 sub-agents, covering architecture, tradeoffs, and what production teams need to know about agent orchestration.
9 min read
Agentic AI & LLMs
Agentic AI & LLMs archive page 270 of 631
Agentic AI & LLMs
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
Latest articles
MCP 1.0 Spec Freeze: What It Means for Tool Builders
Infrastructure-level look at MCP 1.0 spec, including Model Context Protocol, deployment topology, region availability, and cost considerations.
11 min read
Claude Opus 4.7 Tool Use Reliability at 1M Tokens: What We Measured
A practical engineering deep dive into Claude Opus 4.7 tool use, covering architecture, tradeoffs, and what production teams need to know about function calling.
7 min read
Safety Skills: How Constitutional AI 3.0 Surfaces in Production
A practical engineering deep dive into Constitutional AI 3.0 production, covering architecture, tradeoffs, and what production teams need to know about AI guardrails.
6 min read
A2A Protocol: Google's Agent-to-Agent Spec in 2026 Explained
Google's A2A protocol lets agents discover and call other agents across vendor boundaries. The spec, the agent-card flow, and what it competes with in the agent stack.
11 min read
AWS Bedrock AgentCore Deep Dive: Memory, Tools, and Gateways
Bedrock AgentCore adds memory primitives, tool gateways, and runtime services to the Bedrock agent stack. The architecture explained for real production builders.
11 min read
Cohere Rerank v3 vs Voyage Rerank vs Jina: Quality and Cost Tested
Three reranker families dominate 2026 with different pricing and accuracy profiles. Quality on real RAG datasets, latency, and price-per-million ranked pairs measured.
11 min read
Memory-Aware Prompts vs Naive Context Stuffing: Cost Trade-Offs
Stuffing the whole memory buffer into the prompt is a token-bill grenade waiting to explode. The memory-aware prompt patterns that cut cost by 4x in production.
9 min read
10,000 Minutes of Voice AI: CallSphere vs Vapi Cost Comparison
What does 10,000 minutes of voice AI per month really cost on Vapi vs CallSphere? Here is the side-by-side, line by line.
12 min read