AI EngineeringReliable JSON From LLMs: Structured Outputs Compared 2026
Strict structured outputs hold ~99.9% schema compliance while plain JSON mode fails 8-15% of the time. I compare OpenAI, Claude, and Gemini with one Zod schema.
· 7 min read
Articles
Notes from building real things
Web, React and Next.js, AI and RAG, and mobile. Written from production work, not tutorials about tutorials.
FeaturedAI Engineering
How Much Does It Cost to Build a RAG Chatbot in 2026?
An MVP RAG chatbot costs $4k-$12k to build in 2026, production builds run $15k-$40k+, and running costs land at $5-$30 per 1,000 queries. Full breakdown.
· 6 min read
AI Engineering
AI EngineeringDo AI Agents Need a Memory Layer? Mem0 vs Letta vs Zep
Most AI agents don't need a memory vendor. Unless you need consolidation, decay, or cross-agent state, Postgres with pgvector covers memory for $0 extra.
· 8 min read
AI EngineeringHow to Migrate Your MCP Server to the 2026 Stateless Spec
The final MCP spec ships July 28, 2026 and removes sessions from the protocol. I migrated my production Node server; here is the exact diff and checklist.
· 6 min read
AI EngineeringPrivate RAG on Local Models: Qwen3 vs Gemma 4 in 2026
Yes, you can ship private RAG on one 24GB GPU in 2026. I ran a 50-question eval: Gemma 4 26B MoE wins English corpora, Qwen3.6 27B wins multilingual.
· 7 min read
AI EngineeringHow to Secure an MCP Server: 2026 Hardening Checklist
I audited my production MCP stack against the NSA's May 2026 guidance and the OX Security RCE disclosure. Here is the 12-point hardening checklist I use.
· 7 min read
AI EngineeringClaude Code vs Cursor vs Codex for Real Client Work 2026
Pricing converged at $20/$200 and SWE-bench scores sit within a point, so workflow decides. Real cost-per-feature numbers from paid client projects.
· 7 min read
AI EngineeringIs RAG Dead in 2026? Agentic Retrieval in Production
No. I rebuilt my production SaaS pipeline as agentic retrieval: cost per query down 36%, accuracy up from 68% to 89%. Only naive top-k RAG died in 2026.
· 7 min read
AI EngineeringHow to Reduce LLM API Costs: Caching and Routing in 2026
I cut my reputation SaaS's LLM bill 79%, from $41.60 to $8.90 per 1,000 AI replies, using routing, prompt caching, semantic caching, and batching.
· 7 min read
AI EngineeringAI Agent Observability in Node.js with OpenTelemetry
OTel GenAI spans went stable in early 2026. Here is how I instrument a TypeScript agent in Node.js, track cost per trace, and alert on silent failures.
· 7 min read
AI EngineeringHow to Add AI Features to an Existing SaaS Without a Rewrite
You do not need a rewrite to add AI to a SaaS. Add one endpoint beside your existing API, wire it to one workflow, stream the output, and cap token spend.
· 8 min read
Building something similar?
I take on a few projects at a time: web apps, AI features, and mobile. Tell me what you are working on.
Start a conversation