· 8 min read
agentic-ai
6. Make the Chatbot Faster Without Quietly Making It Worse
Optimize context, output, caching, retrieval and inference using an end-to-end latency and token budget.
Read moreIDEAS, DECISIONS & LESSONS
Notes on AI, architecture and building useful products.
51 articles
· 8 min read
agentic-ai
Optimize context, output, caching, retrieval and inference using an end-to-end latency and token budget.
Read more· 10 min read
agentic-ai
Prepare supervised examples, configure an adapter experiment and decide whether fine-tuning earns its operational cost.
Read more· 8 min read
agentic-ai
Build independent evaluation sets, score grounded answers and measure quality, cost and coverage together.
Read more· 8 min read
agentic-ai
Compare practical local model candidates, CPU routing and GPU fine-tuning without confusing size with production quality.
Read more· 10 min read
agentic-ai
Design a governed data agent with a semantic layer, read-only tools, tenant isolation and verifiable answers.
Read more· 8 min read
agentic-ai
Define chatbot jobs, risk boundaries, quality targets and a measured path toward fewer model dependencies.
Read more· 8 min read
agentic-ai
A CPU-only chatbot and agent routing demo: trained classifier heads, bilingual tests, measured LXC latency, failure cases and a live homelab deployment.
Read more· 5 min read
agentic-ai
A tested NanoClaw publishing workflow with draft-first writing, content fingerprints, CMS permissions and explicit review limitations.
Read more· 5 min read
agentic-ai
Anthropic's September Claude Code update quietly raised bashOutputMaxChars and taskOutputMaxChars to 128K and added a file-based subagent system prompt flag. A Tech Lead's read on why this unglamorous change matters more than most model announcements for anyone running fan-out subagent pipelines.
Read moreLooking for an article, exploring a product, or starting an AI project?