Model providers, agent orchestration, MCP, RAG, observability, and the app stack — what we use, and when we use each. Built for technical authority, not affiliate links.
Agent apps, tools, structured outputs, tracing, and handoffs
Codebase work, Claude Code, long-form reasoning, repo-aware engineering
Google ecosystem, enterprise agents, document-heavy governed workflows
Real-time, search-aware systems with web and X search
Durable orchestration — checkpointing, retries, branching, human-in-the-loop
Tools, handoffs, and tracing for multi-step OpenAI agents
Streaming, RSC, and the modern AI app interface on Next.js
Bespoke control loops when off-the-shelf frameworks don't fit
The universal standard for connecting agents to tools and data
Purpose-built servers exposing your tools and workflows to agents
Catalogued, governed tools agents can discover and call
Internal-only tools behind auth, scoped to your systems
Vector storage and similarity search inside Postgres
Postgres, pgvector, and auth — the AI-native backend
Managed vector database for large-scale retrieval
Vector database with hybrid search and filtering
Combining keyword and vector search for better recall
Reordering retrieved passages for relevance before generation
Retrieval architecture over proprietary knowledge
Tracing, prompt management, evals, cost, and latency
Tracing and evaluation for LangChain and LangGraph apps
Built-in tracing for OpenAI agent and tool runs
Bespoke views into cost, latency, and agent behavior
The React framework for production AI apps
The component model the interface is built on
Type-safe end-to-end application code
Data pipelines, evals, and ML tooling
Postgres, pgvector, and auth backend for AI apps
The relational core for application data
Caching, queues, and low-latency state
Edge compute, storage, and delivery
Deploy, preview, and run the Next.js app
Build logs, agentic engineering decisions, agent failures, evals, and what survives real users. Sent weekly, never more.