Issue No. 41

People often ask us why we don't face more competition for the deals we do.

The simplest answer is that our transactions are too small for most large credit funds to pursue economically. A $10 million facility can require just as much underwriting, legal documentation, and oversight as a $50 million facility.

I've always approached business as a builder. Before starting Pier, I'd built businesses from the ground up, including the processes and infrastructure needed to make them work. We brought that same mindset to lending, treating our credit facility offering like a product that we could continuously improve.

We close more than 20 transactions a year, giving us plenty of opportunities to refine our underwriting, documentation, and processes. We've been relentless about incorporating lessons learned into standardized templates, allowing us to become more efficient with each transaction.

We've also designed our facilities around what our customers, typically CEOs and CFOs of smaller specialty finance companies, actually value. One example is our decision to offer draw-down facilities rather than traditional revolvers. Revolvers can be cheaper because borrowers only pay interest on outstanding balances, but managing daily draws, repayments, and reconciliations requires operational resources many smaller companies don't have. By removing the revolving feature, we've made the product easier for borrowers to use while reducing our own risk.

Our goal is to keep getting better at both sides of the business: making more thorough credit decisions in less time and at lower cost, while continually improving the product we offer to smaller specialty finance companies. I think there's still plenty of room to improve on both fronts.

— Jillian

‍

As we continue to build out tools with AI, I have been thinking about the difference between replacing existing tools versus building all new AI-native tools.

What's the difference?

If we vibe code a replacement for our CRM with AI, the new CRM is static as of today. In a year, we need to upgrade it again. But, what if AI is built into the CRM itself? Then, as AI improves, the CRM itself improves. Since AI is improving at an exponential rate, our CRM can improve at that same rate.

But, how do we make it recursive?

Here's another example. Each day, I get a daily brief from AI. It reviews my email, Slack, calendar, Asana, CRM, and several other systems to provide me a 5-minute review of everything important to me for that day. That brief is created off of a scheduled prompt. AI runs that same prompt every day, so my report itself is improving exponentially as the underlying model for AI improves.

But, here's the twist. That brief has a boatload of interactive features. I can draft emails from it, rate parts of the brief, add tasks to Asana, re-order priorities, move calendar invites, click through to emails or links, and more. Once I'm done with the brief, my interaction with it is recorded. Then, each day, the prompt reads through all of my previous interactions with the brief to think about how to better provide the brief today. That is the recursive part. So, now my brief improves itself (recursive) while growing at the pace of AI (exponential).

Now imagine that on a deal, pricing up loan portfolios, or modeling portfolio risk. We're not there on all of these yet, but recursive exponential improvement is in sight in Pier.

- Conor

‍