Workshop: Build Your Macro Research System - NYC Summer 2026

A one-day, in-person, AI working session in partnership with Haver Analytics. Dates and pricing announced soon.

The Premise

Ask any AI tool where US inflation is heading and you'll get a competent summary of consensus — hedged, generic, already priced in. Ask the same question through a system that holds your framework, the research you trust, your data definitions, and your house view, and you get something that looks like the start of an actual note.

The model didn't change. The context did. That gap — between AI as a toy and AI as research infrastructure — is a buildable skill. Not prompt tricks: context engineering, the discipline of designing the information environment your analysis actually runs in. This workshop teaches you to build it, hands-on, in a day.

What You Build

The whole day runs on a single live macro question — the kind you'd be asked the week you attend. Every segment works that question at a deeper level of capability.

Prompting → context engineering → agentic workflows → an MCP server on your own machine

• A working macro research system in Claude — a research analyst built to reason the way you do: your framework, your house view, your standards. You'll test it against the same question put to a naked model, and grade both as the analyst you are. It leaves the room with you.

• The shift from prompting to agentic work — the difference between driving an AI step by step and setting a research goal, supervising the work, and judging the result. The shift that decides whether AI saves you minutes or changes what your desk can produce.

• An MCP server, implemented by you — connecting Claude to live data and tools, running on your own laptop, so you see firsthand how connected models change what a research workflow can be. This is the part no video course can offer, because it happens on your machine, with you doing the work.

• A Monday-morning plan — one real task from your own desk, mapped against the architecture, so you leave knowing exactly where this applies to your work.

Who It's For

People who cover macro for a living — buy-side and sell-side analysts, economists, hedge fund and asset allocation professionals, policy economists, and industry analysts whose work runs through forecasting, markets, and the macro picture. You don't need to be technical. You do need to be serious: this is a working session, not a webinar, and you'll build, run, and judge real systems on your own machine.

For Teams

The open workshop also runs as a private, in-house session — your team, your workflows, and your own data infrastructure, scoped to what your desk actually does. Where the open course works on a shared question, an in-house engagement builds the system on your stack, and can go further into the parts the open room can only see demonstrated.

If you're considering sending several people, the private version is usually the better one to start with. Tell me who the team is and what you want them to be able to do.

Register interest → Dates and pricing are being finalized. Register interest and you'll be first to hear when registration opens. daniel@smartmacrolabs.com