Built This Week/Builds

Builds

Portfolio Macro Scenario Simulator

Sam's demo runs a macro scenario against a 30-company seed portfolio and buckets every company into unaffected, advantage, stress or watch, a feature idea he built for Emblem.

For Episode 31 (February 13, 2026), Sam built a portfolio macro scenario simulator. You give it a portfolio, pick a macro scenario, and AI estimates how each company would fare. It's a demo on sample data, built with our guest's company, Emblem, in mind. The demo itself lasts about two and a half minutes on the show, and Sam didn't say how long it took to build or which tools he used, so this page is short.

Why we built it

Our guest was August Kiles, head of product at Emblem, a platform that indexes fund data rooms and helps VC, growth equity and private equity firms with diligence, reports and financial models. Before the episode, Sam and August talked through possible future features for Emblem. A portfolio scenario tool was one of them, so Sam built a version to show.

The stack

Not covered on the show. Sam presented the finished tool and the "AI is doing its work" step, but didn't name the coding tool or the model behind the simulation. We won't guess.

How we built it, step by step

What we can describe is how the tool works, as Sam walked through it at 02:20:

  1. Load a portfolio. The demo uses 30 made-up seed companies (names like SwiftPay and AgroLogistix), each with its investment round and year.
  2. Pick a macro scenario. For example, an increase in regulatory pressure.
  3. Set the parameters. Speed of change, whether the shift is temporary or lasting, market sentiment, and a time horizon of 6, 12 or 24 months.
  4. Run the AI simulation.
  5. Read the buckets. Companies land in unaffected, advantage zone, stress zone or watch list.
Maybe it's it's only a temporary potential macro scenario happening.

— Sam Nadler, Episode 31

How it turned out

Sam ran two scenarios live: regulatory pressure, and a tongue-in-cheek "liquidity supernova" where every company has a liquidity event. He was upfront that it is "obviously a demo tool," something Emblem might layer on later.

August's take at 05:15 was positive. He said a tool like this could be useful to a large portion of Emblem's potential customers. His bigger point was about speed: funds invest across so many sectors and stages that their workflows vary a lot, and because new features are now so easy to build, Emblem can ship them on demand. He described that as a flywheel for winning markets.

What we'd do differently

We didn't test the simulation's outputs against real outcomes, so there's no evidence yet on how accurate it is. To be useful beyond a demo, it would need a real portfolio and someone checking its calls over time.

For another of our forecasting-style demos, see the Monte Carlo pipeline forecast dashboard. More on the industry: AI in Finance.

FAQ

How does AI portfolio scenario analysis work?

In Sam's demo you load a portfolio (30 seed companies with their round and year), pick a macro scenario such as increased regulatory pressure, set speed of change, whether it is temporary or lasting, market sentiment and a 6, 12 or 24-month horizon, and the AI simulation sorts each company into unaffected, advantage zone, stress zone or watch list.

Can you stress test a VC portfolio with AI?

Sam's tool is a demo on sample companies, not a production system. Emblem's head of product, August Kiles, said something like it could be useful to a large portion of Emblem's potential customers.

What tools were used to build the portfolio scenario simulator?

The episode didn't say which tools or model Sam used. He showed the finished demo and moved on to the interview.