Built This Week/Guests

Guests

August Kiles

August Kiles is head of product at Emblem, an AI platform that indexes investment funds' data rooms so analysts get answers, reports and financial models in seconds and get to a no faster.

August Kiles is head of product at Emblem, an AI platform for investment funds. Emblem indexes a fund's data rooms and powers search, deal comparison, financial extraction, financial models and reports that export to Word, PowerPoint and Excel. He joined us in Episode 31 (February 13, 2026), the week Claude Opus 4.6 had everyone talking.

What Emblem does

August described Emblem as a one-stop shop for anyone working in venture capital, growth equity, private equity or family offices. It tracks a deal through the whole life cycle of a fund, starting the moment a data room arrives.

Emblem's founder and CEO had worked in private equity and interned in venture capital. When LLMs spread in late 2022 and early 2023, he saw right away what they could do for diligence.

The demo

At 04:04, August walked us through Emblem, starting from its landing page line, "the last platform investors will ever need." The flow he described:

  1. Sync the data room. As soon as a data room comes in, you sync it with Emblem and every file gets indexed.
  2. Ask specific questions (07:02). Instead of associates combing through files, you ask something like the 2017 revenue for one product line, without knowing which spreadsheet holds it. The answer comes back in seconds.
  3. Generate reports (08:07). Give Emblem a list of metrics and the analysis you want, and it produces the results in seconds, ready to export.
  4. Route each task to the right model (12:35). Emblem runs a multi-agent setup. Claude and Gemini do most of the work in the product, and Grok sits in a model dropdown for users who prefer it.

Associates still do the last 5%: deciding what to ask Emblem and presenting the results inside their firm.

Key ideas from the conversation

  • Getting to no faster (09:13). The deals a fund does still take as long, but a no can take a few hours instead of a hundred. That means more companies reviewed and more diamonds in the rough.
  • More useful the later the stage (11:10). Seed investing runs on instinct and simpler models. Private equity data rooms have thousands of files and years of complex financials, which is Emblem's ideal case.
  • AI coding went too far, then came back (13:20). Emblem leaned on AI-written code early in 2025, shipped too many bugs and rolled it back over the summer. With the latest models, most of Emblem's code is AI-generated again.
  • The new Opus builds better financial models (13:57). August saw a big jump in generating financial models and outputting them to spreadsheets.
But if you can say no in, you know, say just a few hours of work instead of a hundred hours of work, then you can look at that many more companies.

— August Kiles, Episode 31

The model matters more than the tool

In the news segment on Amazon engineers wanting Claude Code, August called Opus 4.6 "truly an inflection point." With good architecture and prompting, he rarely has to adjust what the model writes. His contrarian take was that people bandwagon on tools.

At the end of the day, everything out there is really good and using, you know, say Cursor or Copilot with Opus 4.6 is going to be so much better than using Claude Code with Opus 4.5 that it's not even funny.

— August Kiles, Episode 31

He still uses Cursor for its in-editor browser and debug mode. Sam agreed the model mattered more than the IDE. For our own take, see Claude Code vs Cursor.

What we built for the episode

Sam built a portfolio macro scenario simulator with Emblem in mind. It runs a 30-company seed portfolio through a scenario like rising regulatory pressure and sorts each company into advantage, stress or watch zones. August said it could be useful to a large portion of Emblem's potential customers. More on this space: AI in finance.

FAQ

Who is August Kiles?

August Kiles is head of product at Emblem, a startup that optimizes workflows for investment funds. At Emblem's early stage he also writes a lot of the code himself.

What does Emblem do?

Emblem syncs and indexes a fund's data rooms, then offers search across single and multiple deals, financial data extraction, financial model building and reports exported to Word, PowerPoint and Excel. It targets VC, growth equity, private equity and family offices.

Will AI replace private equity analysts?

August thinks not. He says Emblem frees analysts to do more and helps funds get to a no faster, so they can look at more companies; junior salaries don't move fund returns, but finding the right deals does.

Which AI models does Emblem use?

Emblem uses a multi-agent setup that routes each task to the best model. August said Claude and Gemini are used most in the product, with Grok and OpenAI models also available.